Only One Science: Twelfth Annual Report of the Et National Science Board

Home , Binary code, Iced Out Audemars, Z5 (computer)

DOCUMENT RESUME_

ED 215 889 SE 007 086 . TITLE Only One Science: Twelfth Annual Report of the et National Science Board. INSTITUTION National Science Foundation, Washington, D.C.?- National Science Board. REPORT NO NSB-80-1 PUB DATE [81] - . NOTE , 221p.. AVAILABLEFRIOM Superintendent of Documents, U.S. Government Printing Office, Washington',, D.C. 20402.

4 EDRS PRICE' MF01/PC09 Plus Postage. DESCRIPTORS Computers; Environmental Education; Federal Programs; *Medical Research) Medicine; $,Natural Sciences; Pesticides;' Radiology; Science Education;.*Scientific Enterprise; *Scientific Research; Seismology; Semiconductor Devices; *Social Sciences; Surveys; Technology IDENTIFIERS 1,1tional Science. Foundation; *Science and Society ABSTRACT Departing markedly from previous reports to Congress by the National Science Zoard, this document presents in an infoxrdhl, narrative style six, stories depicting scientifit discoveries and their effects on society,. Drawn from the'phyiical, biological, medical, and-social sciences, topics discussed include: (1) computers and Semiconductors; ,(2).seksmic exploration for oil and' gas;(3) survey rtAgarch and opinsin polls; (4) pesticides and pest control; {5) synthetic fibers; and (6) x-rays for medical diagnosis., Each chapter examines the technological, social; and political aspects of 'these current issues and illustrates hl)w major scientific developments most often emerge from the combined efforts of individuals, universities, industrie, and government. Illustrations and photographs complement the text. The'publication is not iAtended as a policy or budgetary gUideNbut as an example of the interrelationship between-science and the general welfare of society. (DC)

****************************************A****************************** Reproductions supplied by EDRS are the best that can be made * from the original document. **************4******************************************************** TWELFTH 1)ONLYONESCIEN E ANNUAL REPORT OF THE NATIONAL SCIENCE BOARD rr

US DEPARTMENT OF ENCATION NATIONAL INSTITUTE OF EDUCATION EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC) \/This document has been repruced as Et received from the person or organization originating Minor changes have been made to improve reproduction quality

Points of view or opintons stated ,n OuS.O0Cu ment do not necessaniy represent official NIE POS.b0o or pOcy

"PERMISSION TO REPRODUCE THIS MATERIAL HAS BEEN GRANTER BY

1b0 01)kri ora

TO THE EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC)"

NATIONAL SCIENCE FOUNDATION MEMBERS OF THE NATIONAL SCIENCE BOARD

DR LEVI,S'M BRANSCOMB (Chairman, Na- DR. DAVID V. RAGONE, President, Case Western tional Science Board), Vice President and Chief Reseriie University, Cleveland, Ohio Scientist, International Business Machines, Inc, DR DONALD B RICE, JR President, The Rand Armonk, New York Corporation, Santa Monica, California MR HERBERT D DOAN (Vice Chairman, Na- DR STUART A RICE, Frank P. Nixon Dft, tional Scienee....Board), Choirtfian, Doan Resources tinguished Service Professor of Chemistry, The Corporation, Midland, Michigan James Franck Institute, University of Chicago, ,DR RAYMOND L BISPLINGHOFF, Vice Presi- Chicago, Illinois dent for Res'earch.and Development, Tyco Lab- DR ALEXANDER RICH, Sedgwick Professor oratories, Inc , Exeter, New Hampshire of Biophysics, Department of Biology, Massa- DR LLOYD M COOKE (Vice Chairman, Com- chusetts Institute of Technolo, Cambridge, mittee on Twelfth NSB Report), President, Massachusetts National Action Coyncil for Minoritiesin En- DR EDWIN E SALPETER, J G White Profes- gineering, Inc , Nei./ York, New York sor of Physical Sciences, Cornell Unirsity, DR EUGENE H COTA-ROBLES, Pirofessor of Ithaca, New York Biology, Biology Board of Studies, University DR CHARLES P SLICHTER, Professor of Physics of California at Santa QuzSanta Cruz, California and in the Center for Advanced Study, Univer- DR PETER T FLAWN, President, University sity of Illinois, Urbana, Illinois of Texas at Austin, Drawer T, Austin, Texas Ex Of hoc, DR ERNESTINE FR1EDL, Dean of Arts and Sci- ences and Trinity Colli:ge, Professor of An- DR JOHN B SLAUGHTER, Director, National thropology, Duke University, DurhariClorth Science Foundation, Washington, D C Carolina Executive Secretary DR MARY L GOOD, Vice President, Director of Research, UOP, Inc, Des Plaines, Illinois MISS VERNICE ANDERSON, National Science DR JOHtI R HOGNESS (Chairman, Committee Board, Washington, D C on Twelfth NSB Report), President, Associa- tion of Academic Health Centers, Washngton, IMMEDIATE PAST MEMBERS OF D C THE NATIONAL SCIENCE BOARD DR WILLIAM F HUEG, JR, Professortif Agron- omy and beputy Vice President and Dean, In- DR RICHARD C ATKINSON (Past Member ex stitute of Agriculture, Forestry, and Home Eco- officio), Chancellor, University of California at nomics, University of Minnesota, St, Paul, San Diego, La Jolla, California Minnesota DR JEWEL PLI4MMER COBB, President, Cali- DR ,MICHAEL KASHA, Distinguished Profes- fornia State University atullerton, Fullerton, sor of Physical Chemistry, Institute of olec- Caljornia- ular Biophysics, Florida St e versity, DR. NORMAN HAOKERMAN (Past Chairman, Tallahassee, Florida ' 1 National Science Board), President, Rice Uni- .00 DR MARIAN E KOSHLAND, Professor of versity, Houston, Texas Bacteriology and ImMunology, University of DR W N HUBBARD, JR, President, The Upjohn California at Berkeley, Berkeley, California Company, Kalamazoo, Michigan DR PETER D LAX, Professof of Mathematics, DR...SAUNDERS MAC LANE, Max Mason Dis- 'Courant Institute of Mathemat.ical Sciences, tirtguished Service Professor of Mathematics, t'Iew York University, New York, NON York Department of Mathematics, University of. DR WALTER E MASSEY, Dire'ctor, Artonne Chicfago, Illinois National Laboratory, Argonne, Illinois DR GLOVER E. MURRAY (Past Vice Chairman, 'DR HOMER A NEAL, Proyost, State University National Science Board), University Professor, of New York at Stonybrodk, New York Texas Tech University and Texas Tech Univer- DR. MARY JANE OSBORN, Professor and Head, sity School of Medicine, Lubbock, Texas Department of-Microbiology, University of DR L. DONALD SHIELDS, President, Southern Connecticut School of Medicine, Farmington, Methodist University, Dallas, Texas Connecticut DR JAMES H ZUMI3ERGE, President, Univer- DR JOSEPH M, PETTIT, President, Georgia In sity of Southern California, Los Angeles, Cali- stitute of TechnOlogy,,Allanta, Georgia forma -

s i

ONLY ONE SCIENCE TWELFTH ANNUAL REPORT OF THE NATIONAL SCIENCE BOARD i \ ,., '..

,..

To him who devotes his life to science, nothing can give more happiness thancreasing the numJer of discover&s./ But his cup of joy is full when the results of his studies immediately find practical application.

There are not two sciences, There is ONLY ONE SCIENCE , and the application o f science, and these two activitiesare linked cias the fruit is to the tree. Louis Pasteur

r ) r i

NATIONAL SCIENCE FOUNDATION

4 Letter of Trarismittals

September 28, 1981 My Dear Mr. President: I have the honor of transmitting toyou and through you to lr the Congress the Twelfth. Annual Report.of the National Science Board Only One Science. As the title implies, thcis Report is a departure from previous ones. In a narrative form it tells about scientific research and describes how the results of such research affect and. benefit society. The Report does not make specific policyor budget recommendations. The introduction to the Report quotes Louis Pasteur: "There is Only One Science and the application of science, and these two activities are linked as the fruit is to the tree.'The Board believes that this is true. We also believe that it is of great importance for all Americans to appreciate how research, technological development, and human welfareare inevitably and necessarily interrelated and intertwined. These stories should help in achieving that understanding. In preparing the Report, the Board selectedas subjects six fields of scientific endeavors froma long list of potential choices. Some, if not all, of these stories should appeal to readers with different interests. All of the chapterswhose subjects range fiorn hc,4, the seismic system is used in exploring for gas and oil to the uses of X rays in medical diagnosisare of current interest. The enabling legislation which created the National Science Foundation Mandates the Foundation "To promote the progress of science; to advance the national health,prosperity,,, and welfare, to secure the national defense and other purposes.' In an informal way this Report deals with the relationship of science to the "general welfare." We hope the storiesare interestingthe record of science stands for itself. Respectfully yours

a-y142,f-047,44- Lewis Branscomb Chairman, National Science Board The Honorable The Resident of the United States ui ... Preface ,

Each ydr the National Science-Board submitsto the President and through, him to the Congressa report which deals with issues of concern to the Board and to the National Science Foundation. Every other year the report is entitled Science Indicators;it consists of an up- dated series of data and indices whichportray the status of science and technology in the United Slates. Inintervening years, reports of' the NatiOnal Science Board have dealt witha wide range of 'scientific issues which the Board feelsare 'important. The Twelfth Board Report representsa significapt departure from previous reports in that it attempts, ina narrative fashion, to deal with a somewhat broader aspeCt of the NationalScience Foundation's responsibilities establishingthe Foundation in 1950, theCongress, among other th.s, directed the National ScienceFoundation and the BOard to "appraise the impact ofresearch upon industrial development and upon the general welfare." ThisReport was developed to do that through the.use, of historical, anecdotalstories of discoveries in six different representative fields ofscience. Although the subject of_eaclihapter is different,. the story each tells is characterized by change:To look atany of the fields at a given moment in the past andto trylo prdject howa thread of ideas, activities, kind circumstances would unfoldin the future, would be difficult, if not impossible. Howewer,the perspectives gainedfrom The Report should give readers a keener sensebf theinteractions among scientific research, induistry, academia,.-andthe individual. The Report shouldnot be viewed as a comprehensivestatement of the current status of a rarticular field oras a basisfbi future public policyexcept to the general extent, that "past isrologue" to the future. - tn the course of rreparan of the Report, various aspects of it were reviewed by thecurrent Members of the National Science Board anikby N .

_.. , the immediate past Board Members, all of whom are listedinside t front cover of this Report. The Committee responsible for the preparation for the Twelfth Boa,rd Report wishes to acknowledge the extensive participation of all present and immediate past Members of the Board and wishes particularly to express appreciation to a' b Dr Gwynn C. Akin, Staff Director for,the Report and Consultant to the .National Science Board, who was the principal staff architect of the Report. The Committee also wishes to express thanks to Dr. Carlos Kruytbosch, the Executive Secretary to the Comthittee and to Dr. Allen Shinn who served in that same capacity in the early months. In addition, more than 250 individuals provided information, documentation, written text, advice, suggestions, c itical comments, or technical assistance during the course of prepara ion of the Report. "------.., Their names are listed in the Acknowledgment section in the back of this 41ume. If any one has been omitted, we offer our sincere apologies COMMITTEE ON THE TWELFTH BOARD REPORT ,., John R Hognest Chairman Lloyd M Cooke, Vice Chairman Herbert D Doan ,,,,..... Walter E Massey Joseph M. Pettit

N..

VI 7

4 J . Contents

...

Introduction 1 r Computersand Semiconductors 7 Seismic Exploration for Oil and Gas . 39 Survey Research and Opinion Polls 8.1 Pesticides and Pest Control 117 Synthetic Fibers 51 X Rays for Medical Diagnosis 179 Acknowledgements 215

. ..

vii EntroduCtion

A Jo him who devotes his life to science, nothing can give more happiness than increasing the number of discoveries But his t upfifroy is full when the results of his studies immediately find practical application.

jhere are not two gciences There is ONLY ONE SCIENCE and the application of.science, and these two activities Iare linked as the fruit is to the tree. eouis Pasteur

No age in history has come closerto providing Pasteur's link than the twentieth century. In a world of explosive changes, the extraordi- nary growth of science and technology has had an impact on everyone and on every aspect of life. ,The'results of scientific advances automobiles, airplanes, wash-and-wear clothes,- antibiotics, television,, glass windowpanes, air conditionersare pervasive and ubiquitous. And yet they are often taken for granted or go completely unnoticed. Dealirtg with the rapid rate of scientific and technological change and with its results presents a unique challengea challenge which can be met only by strengthening the bridge of understanding between the individual and the world of science. The public understanding of the purposes and effects of science and technology is essential to the health and vitality of a modern society. This realization has led the National Science Board to take a novel approach in its Twelfth Report. Rather than presentkg a foimal glide to policy, the Report describa, in informal, narrative style, how certain scientific discoveries occurreasasnd how they have affFcted society. The Report examines six topics of importance in which research, technological innovation, public need,(and human welfare have, Aver varying

1 INTRODUCTION L

periods of time, cone together to createa body of scientific knowledge and a related technology that enhance the quality of life. The Report is meant to show that, in the wordi of JulesHeriri Poincare, "Science is built up with facts,as ihouseis with stones. But a collection of facts is riQ more a science than a heap of stones isa house." Even the failures,,thatare inevitable in research' provide new insights. Thecontamination ofa culture plate resulted in the finding of the antibacteripl properties-of penicillin;a disproven hypothesis call lead to a Gesh view of a problem under study. The six topics examined (computers and semiconductors, pesticides and pest control, tl1seismic systemsurvey research andopinion polls, synthetic fibers, and,X rays for medical diagnosis)ref5r.iserjt a spectrum of scientific endeavors in the physicalbiological, medical, and social sciences. The topics were selected froma group which met several criteria: a signifi5,ant impact ortsociety and the individual,a substantialinformati,p-A base,a history sufficient to assess their developmeht, a firm relationship to scientific research,and general current interest. Each chapter illustrates the problems of dealing with the technological, social, and political realities of society anddemon- strates the strong recurring links among accomplishments by individuals, universities, industriesreerrd the government that bring_about the introduction to society of almost every major scientific develop- I ment. One striking theme that recurs throughout these chapters is that approachesto solutions of scientific problemsare diverse and varied. The popular concept of scientific and technological development is 4 that of a "clean" linear progression from basic, "nontargeted" research leading to applied research which, in turn, leads to technological development-and the marketing of a product. This orderly progression occurs only rarely. When this direct linear relationship' between a basic observation and the practical application does exist, the copnection may be made within a matter of months or yearsor the funda,mental observation may lie .Normant for centuries before itsuses can be perceived apd a new technology developea. Alternatively,a breakthrough in technology may stimulate basic research, which inturn allows tiv investigator to delve deeper into nature's secrets, thus genei.atingmoire, knowledge that can be applied to improve the technology. Ora peed for a particular technology may be perceived, although the basic knowledge necessary for iM development is not yet available. The, need for the technology may stimulate fundamental research which permits the technological development whichmay, in turn, provoke additional basic research.

2 ONLY ONE SCIENCE ,

c Although basic research often leads to technological application and technological ahlication fosters basic research, serendipity is also an ingredient of research. It is te "X" factor that is both.unpredictable 1 I. 4. and often integral to discovery. 14 is the chance Observation that stimu- lates a newidea or a new scient is application. But here, too,new ideas will occur as a result of such a circumstance only'whena background .. of knowledge is already present. For, as Pasteur observed, "In the fields/ of observation, chance favors only the mind4. that is prepared." One of the early examples of the use of seismic wavescame fOrn. a military officer's trying to discover if the enemywere digging a secret tunnel. The officer's basic tools were apan of water, set flat on the ground, and a sharp-eyed soldier stationed to look for ripples in the water caLvecl by digging. Today, the interlocking combinations of basic research, chance observation, need,,and tefhnological development have replaced that soldier with a Host of trainedgeop .y- ,.,0 sicists and that pan of water with sophisticated recording instruments About the only thing that has not changed since that soldier's lonely vigil is a strong urge, sometimes even a need, to know what is goingon Linder the earth's surface Whether it is the search fornewergy

k. resources or for a better way to predict earthquakes, additi nal bask 'research and new technology keep-driVing each other to fu.11 an age- old quelfor answers. \ t Recently, from data developed using the seismograph and other technologies,a new scientific theory was borhplate tectonics ., Accordihg to this concept the crust of the earth is comprised of about 20 plates that are in continual motion. (Plates are huge blocks of the earth's crust that float on the dense, hotter,more fluid rock below ) The kribwledge gained from studying this new field has,, in turn,pro- . vided a greatly improved underStanding of the movement of thecon- tinents, volcanic activity, and earthquakes. Perhaps this new concept will not provide miich practical application for marfy years. But then it' is unlikely that the person who conceived the first abacusmore than 2,000 years ago could have dreamed of a pocket-size calculator cbm: ''i plete with 30 mathematical functionS and a memorycore.' It was not until this Report. was well underway that members of the Board became !ware of another type of interepting interrelationship among the topics. There are many eross-connectiobetween the developments in the various areas of study, a technolog discussed in one of the chapters often has significant applications tthe technologies discussed in ottiers. FOf example, although the dicussion of X rays deals mainly with medical diagnosis, radiation isa so used in / ,test control to sterilize male insects or to develop mutatis in seeds so, t

3 INTRODUCTION '

s e , 1..1. ek

7 that pest-resistant crops can be.grown. A clear relationship also exists between the seismic system and synthetic fibers-Toil and gas discovered by using that system provide the base substances for the production of synthetic fibers. The most impressive interconnection, however, is the relationship.of the computer to thebther subjects discussed. For example, computers are an integral part of the developinent of the CT (computerized tomography) scanner so importanin radiolOgical diagnosis today. Many advances in theuse of the seismic system to help locate oil and gas could not hive occuDred without computers. Computers are used in analyzing data from large scale statisti,- "cal surveys. They are even used in the control of production processes in the manufIcture of synthetic fibers and in numerous calculations neeessazy for the success of certain pest control techniques. One technological development often makes another developmerlt, possible, and so science builds upon science, and one technology pn another. Discovery is a product of opportunity, imagination, brilliance, persistence, and serendipity. Unfortunately, none of these factors can be mciasured exactly, nor can they-be ordered at will. An enlightened society must Tecognize the need for major opportunities in unfettered research in wRiCh imaginative scientists feel free to pursue their curiosity beyond the limits of current knowledge. To extend Pasteur's analogy of the fruit and the tree, if society expects a bountiful harvest, it Tust constantly nurture and feed the "tree" of science. It must also remember that no one can predict exactly what or when this particular tree will bearonly that, in time, it will.

1 4 G41LY ONE SCIENCES

_a -1

--s

...... ,

. .:, . .0 .4.

# ,,

...... - C . -.. 4

0I 1 t.- V/ t . if et \

----Oil). -I# ( ,Computers and .7 4SO

0 SemiQonduotors ( ..

lt r A fesummers ago, the 22-gar-old victim or an automobile # accident Was rushed to Methodist Hospital of Indiana with critical head injuries. During the next 21/2 days, while the patient's respiratory and circultrfory systems were maintained by machines, doctors ran a variety of sophisticated tests but failed to fin even a glimmer of neurological activity. The patient remained inc a. Finally, \doctors concludedThat he had suffered total brain dea 7Because the victim had been in otrerwisxcellent health, the t doctors brought up the possibility of using his kidneys for transplant ,-; purposes. The next-of-kin agreed to the donationthe process of finding suitable recipients rbegan. Doctors fyicl to find two people who were in critical need of a,kidney transplant and whose immune systems were compatible with that of the donor. If the blood and tissue antigens of the dorior and receiver are not the same or very similarand there are tens-og-thousands of $ possible combinationschances areioverwhelming that the recipient's immuno-defense mechanism will attack, and probably reject, the .. . donor's kidriey. In an effort to find a sufficiently close match,lhe donor's types were keyed into a terminal at the hospital and transmitted to a computer operated by the South-Eastern Organ Procurement Foundation JEOPF) in Richmond, Virginia. There, a-10-second search of almost 7 1;500 possible candidates registered by 40 transplant centers in a 17- state area Rroduced a printout of more than 100 names. A silicon wafer (in the fore-.0. ground) is being loaded into a Although all the people on the list had the same blood type and ay with a vaLuurn penal A one or more matching tissue antigens !only two of them met enough d fusionfurnaceused in trite- gr ted eireuxifabrication is in transplant criteria to be considered as prime Prospects: a 34-year41c1 the ba ruund - Indianapolis housewife who had been receiving dialysis treatments for Natmil 5emn.orhii.tor Int the past 7 years and a 40- year -old businessman in Pittsburgh, pennsyl-

:4", 7 COMPUTERS AND SEMICONDUCTORS

14 vania, who was failing on his dialysis treatments and was in desperate need of a transplant. As soon as the SEOPF computer relayed this information to India: napolis and to the transplant centers where the recipientswere registered, 4 surgical team removed the accident victim's kidneys, flushed them with a preserving solution, and packed them in ice-filled containers. One was raced across town in an ambulance, the otherwas hand-carried to Pittsburgh on a commercial jet. Less than 8 hours after the kidneys had been removed from the victim,a message was flashed by the computer to all terminals in the SEOPF network: Two, kidneys transplanted with successful results. While transplant-kidney-matching isone of the more perigheral computer applications, it does illustrate how far the state of the art has come in just 30 year. The early vacuum-tube computers of the late /1,9408 and early 1950s were monstrous machines thatwere clearly the .--exclusive province of a handful-of mathematicians, physicists, engineers, and astronomers. Even the,next two generations of data processing machines, built and marketed during the 1960s, could be used. efficiently only by people with a strong background in mathematics and science. Today, however, more and more computer capability is found in the hands of people whose prime interests are in other areas. Computers are used routinely in medicine, in schools, in factorils, in offics)s, andin the home. And while these machines have not yetkecome everyone's tool, there isno doubt that the current generation OcomputerscompaCt, solid-state systems relying heavily on microtechnologyis far faster, more versatile, lessexpen- sive, and easier to use than its predecessors.

THE ORIGINS OF DATA PROCESSING Many people outside the electronics field tend to think of.the computer as a magical engine which sprang into existence full-blown, like Athena from the head of Zeus. In fact, thecomputer has roots that reach far back in historymore than 2,000 yearsto the abacus. Com- puter technology rests upon knowledge accumulated over manycen- turies and evolved from the genius of many inventive minds. No single breakthrough or classic experiment broht this device into being, instead, historians trace the development of computers along at' least four technical tributaries which merged about 40years ago to form one powerful stream.

8 ONLY ONE SCIENCE

15 Automatic Regulators One tributary, often overlooked irirecounting the,background of computers, goes back to James Watt and his steam engine. In 1788 when Watt applied a centrifugal fly-ball governor to his engine, he did much more than just improve the machine's efficiency. In effect,he showed for the firstlime how..a machine could "examine" itsown output and use the information to monitor and control its internal operation . Other automatic regulators followed. In 1830 Andrew..Ike, invented the thermostatto help control the temperature of furnaces, in 1852 Leon Foucault devised thegyroscope, first used to maintain the course of torpedoes, but later to become the mainstay of na ion for ships and airplanes In the 1860s James Clerk Max supplied a mathematical theory which, among other things, helped to establish the science of automatic controls. Today there are feedback mechanismson everything from dishwashers to data processing machines. In concept, however, they all go back to Watt's fly-hall governor, a system of revolving weights that irl1-tQi lathes Aatt desqvied act as an automatic throttle Interestingly, Watt was not searching for' '!, erriL r ,tcit'tatt an abstract principle when he'built the control mechanism. Rather, he the steam ethwre and. to anou, it t? #urIL Non In- was looking for a practital answer to an urgent needa device that ch volLitht:, ,i/aIgt's would enable his steam engine to ma stain operating speed despite Changes in load.

Boolean Logic Another stream which contributed to modern-day data processing techniques flowQd from quite a different direction. Obviously,a computer does not depend onIfly-ball governor, thermostat, or gyroscope to self-direct its activities. But it (nes rely on something equally ingeniousa set of instructions; or a program, which governs the path of electronic signals through the machine's switching circuitry. Arid this ability to process a sequence of logical statements goes back to the work of a remarkable nineteenth-century figure named George Boole. to 1854 Boole published An Investigation of the Lawi of Thought which did for logic what Euclid had done for geometry. The book described c' how logical statemeritscould be translated into precise mathematical forms. Boole's mathematical logic was binary in nature because it was based on-the premise -that statements are either true or false. The switching circuits of digital computers are also binary in nature because they can existIn onlybr of two states: "on" or "off." What Boole d(of the development of computers was to set the stage for the stord-erogram computer:In the earliest data processing machines, circuit patterns were more or less fixed to performa specific

9 COMPUTERS AND SEMICONDUCTORS

FFt s

16 job. The only way a program could be chariged, forexample, was literally tag° in and reset the In 1938, however, mathematician Claude Shannonikof the Massachusetts Institute ofTechnology (MIT) suggested in a technicalpaper that Boolean algebra could be used for more flexible programming of electronic calculators. Eightyears after that, the famed mathematicianJblin von Neumannactually Showed how programs could be written and stored in themachine. . "This resulted in two very great gains,", explainsChristopher Evans in his go(ft, The Micro Millennium. In the first place, one could take advantage of the computer's huge processing speed and allow itto change programs when required; it could switch from one program to another in a fraction of a second instead of relying on the lumbering skills of its attendant human being. In the second- place, and this it far and away the most important point, it meant that programs within the system could interlock and interact... In principle, programs could evenmodify otherprograms, rewriting them to fit the needs of the moment and integrating them with yet others within the suite. Thus, in one conceptual jump, the feedbackconcept that begarr ./with Wag was joined with the mathematical logic of Boole.

Calculating Machines A more direct tributaryone that has been recountedoften in the history of dattprocessingsprang from the abacus.This little device came into being afore than 2,000 years ago and still is the most widely used calculating tool on earth. Invast areas of Asia, in fact, it is virtually the only known counting machine.

The abacus is a device for making arithmetical calculations Similar counting tools have been used throughout ASia and the Middle East for more than 2,000 years S,nahsonwn

The abacus was developed asa direct result of early efforts to count. Probably the first quantitative symbols usedwere a "two" and

10 ONLY ONE SCIENCE

4 17, 4 a "five": "two" because people have two hands and "five" becatise-of the five digits on each hand. It is not surprising that the most popular 3 - form of abacus makes use of the two -five or binquinary notationsys- 4 tem. The Chinese suan-pan, for example,\cofisists of a series of rods on which beads are strung.There are seven beads on each rodseparated by a divider strip into five on one side (each bead equivalent to 1) and two on, the other (each bead equivalent to 5). Thus, the number "seven" can beexpressed on a line by moving over one bead from the "five" side and two beads from the "one" side. 4 In the hands of a skilled practitioner, the suan-pan is an amazingly fast, accurate, and versatile ice. Nevertheless, it has a distinct shortcoming. It cannot ca over tens from one line to another, anti as mathematical horizons wer expanded, this deficiency becamea major problem. Although many counting machineiwere devised over theyears, it was not until the seventeenth century that the next clear-cut advance came along. In 1642 a young Frenchman named Blaise Pascal, working in his father's tax office in.Rouen, built an ornate shoebox-size device which employed linked geafs to add, subtract, and, most importantly, carry over tens. Thirty years later Gottfried Leibniz improved on Pascal's machine so it could multiply, divide, and calculatesquare roots. In effect, these two inventors brought about the calculating age. (Pascal's contribution to survey research is mentioned in the chapter, "Survey Research and Opinion Polls.") During thenext one-half century many new calculating machines were built, but'all were primarily modifications-or refinements of .4:7'ascal's original design. In fact, although thenext important development had nothing to do with calculating, it was vital to advancement 4 of the science. In 1780 Joseph Marie Jacquard builtan automatic weaving loom which operated from instructions punched into cardsor paper tape. This invention, which revolutionized the weaving industry, led directly into one of the most unusual stories in the history ofcomputers. The year was 1822 and the principal was a complexyoung Englishman named Charles Babbage. Babbage, a mathematician and inventor of the railroad cowcatcher and the first tachometer, was becoming increasingly incensed by the many errors he found in insurance records, logarithm tables, and other lists of data. His fetish Charles Babbage for accuracy was so great; in fact, that after reading Lord Tennyson's (1'91-18"1) famed line, "Every moment dies a man/Every momentone is born," he /8V1 wrote to the poet, "It must be manifest that if this were true, the popu-

11 COMPUTERS AND SEMICONDUCTORS lation of the world would be at a standstill!' Babbage's recommended change to Tennyson, "Every momen \diesa man/Every moment 1.1116 a is born." In 1822 Babbage began work on a machine, which he called the "Difference Engine," that could help solve polynomial equations to six ..s places. The Englishgovernment wasso impressed by the machine's potential for compiling navigational and artillery tables that they sub- sidized him heavily. The projected machineit said Babbage, would be able to do complex calculations and print out its results. Itwas to have a "memory" section made up of the same sort of punched cards used by Jacquard'sioom; cards were also to have been used for input to the Y machine and control of its successive operations. The device was-to have an arithmetic unit, called a mill, in which to store data; itwas to be able to set up its own results in type, thus avoiding transcription errors. Babbagewas literally 130 years ahead of himself. He built a small working model, but never was able to complete a full-sized Difference Engine. The reason for his failure, however, had nothing to do with the concept; the machine could not be built primarily because the technology of the time was not adequate to permit construction of the needed parts. By the time the eccentric genius died in 1871, he had managed to put together only a few parts. Nevertheless, his elaborate drawings of the machine left no doubt that he well on his way to a true compttter. This replica of Babbage's Dif- Several years ago, B. V. Bowden, writing in Think magazine, ference Engine was constructed affording to his plans Unlike a described his efforts to track down the Babbage story. Aside from the true computer Babliagesinren- discovery of many papers which proved the genius of the farsighted tnon was capable, of Making complicated computations, but Englishman,the trail led to Lady Lovelace who was the dayighter of the i t had no memory as a true com- famed poet Lord ,Byron. It seems thattady Lovelace was as much a puter does prodigy in mathematics as her father was a master of poetry. She de- IBM Archmes vised, among other things, a form of binarraritlunetic and an "infal- lible" system for prediding horse race winners. The binary system, originally described by Boole, has stood the test of time and is today used in electronic digital computers. The betting system cleaned out the family fortune and forted Lady Lovelace to pawn all her jewels.,

12 ONLY ONE SCIENCE

19 4 ,Somewhere along the line Lady Lovelace met Charles Babbage. Apparently their mutual interest in mathematics was the spark for a long-ilasting relationship. As Bowden wrote: Lady Lovelace often visited Babbage while he was Making his machines, and he would explain to her how they were constructed and used. As one of her contem- poraries recalled, "While the rest of the party gazed at this beautiful instrument with the same sort of expression that some savages are said to have shown on first seeing a looking glass or hearing a gun, [she] understood its working and saw the great beauty of the invention." She worked out some vary complicated programs and would have been able to use any of the modern machines. She wrote sketches for several papers, but publi%hed only her notes on Babbage, and they were anonymous. Perhaps the most' perceptive observation made by La ly Lovelace was one which could just as easily be applied to the present eration of powerful microminiature computers. It concerned the qu of whether Babbage's machine could be considered "cre;tive." S rote, "The Difference Engine has no pretensions whatever to originate anything. It can4do whatever weknow how to orderit to perform. It can follow analysis; but it has no power of anticipating any analytical relations or truths. Its province is to assist us in making available what we are already acquainted with."

Electric Tabulating Babbage, Boole, Lady Lovelace, Jacquard, Pascal, and many others Machines brought technology to the brink of modem data processing. But it took the work of ,one man to push it over into the practical devices &hat are found today throughout the world. He was Herman Hollerith, a statistician from Buffalo, New York, who arrived on the scene just as the United Sates government was about to stagger its way through the once-a-decade census count. The basic problem was simple. the government had needed 7 full tears of counting and tabulating to complete thg 1880census. All data were handwritten on cards, the cards were manually sorted into van- ous categories, each category was manually counted, amt then the cards were resorted into new piles and counted once again. Compiling the census was such a large, tedious job that by the time the final count was completed, it was already outdated. With the time for the 1890 Herman Hollerith census count approaching and immigration swelling population ranks (1860-1929,j by the day, the Census Bureau'cOuldnvision 9 or even 10 full years ko !8M Archives take the next count.

13 COMPUTERS AND SEMICONDUCTORS

20 Alth9ugh only 44.years had elapsed between Babbage's frustrating efforts to build a Difference Engine and the development of the Census Bureau's forebodings over the 1890 census, two things had happened in the meantime to change the technological clirm4e. First, both machining and manufacturing ski 1Whad improved tremendously as a result of the Industrial Revolution; and, second, an exciting new form of povt, electricity, was now being used to drivean increasing number of machines.

4 -

124.7711:33_1,4:41'1127:

Herman Hollerith s tabulator which he developed for use M the United States Census of 1890 ZwItt%,,,tdor

Hollerith worked out an electromechanical method of recording, tabulating, and organizing census data. His system used cards, about the size of the old United States dollar bill, in which datawere recorded in the form,of holes made with a condinor's hand punch. The punched cards wefe then-positioned one-by-oneover mercury-filled cuss in a special machine. At the touch ofa lever, rows of telescoping pins descended on the cards; where there was a hole the pin simply dropped through into the mercury, thusompletingan electrical circuit. The electrical impulse, inturn, was used to movea pointer one position on a dial. As thedials went around, various totalswere

14 ONLY ONE SCIENCE

21 1 2

accumulated (Hollerith's work'vas also a key contribution to the field of survey research and is referred to in the chapteron that subject.) While this method sounds primitive by today's standards, it enabled the government to complete the 1890census of 62 million ,people in just one-third the time of the 1880census of 50 million. When news of t)e system reached industry, an almost immediate demand developed. Before 19070 Hollerith was marketing his unit record machines to many of the nation's largest firms. The New York Central Railroad used them for car accounting; the Marshall Field Department Store installed electromechanical tabulators for sales analysis work, the Penn Steel COmpany in Philadelphia used them ford cost accounting, and the Western Electric Corporation installed several of the machines for sales analysis Meanwhile, Hollerith went to Czar- ist up a similar system for that country's census count Adoption of Hollerith's machines took place at a time when the United States,was embarking on an unprecedented technological surge. From about 1880 to 1930 commerce in America expanded at a prodigious rate. The railroads pushed west, north, and south toopen new markets and create new industries.'Manufacturing firms adopted mass production techniques which helped to increase their productivity many times over The concept of interchangeable parts ensured that a product manufactured or purcVased in one part of the country could be serviced in another. American ingenuity producedmore than 1 3 million new patents in the first third of thelwentieth century alone. In many respects, this was an ideal environment for the introduc-.- tion of data processing techniques Every, bit of this industrial and commercial development created accounting and recordkeef)ing problems In 1911 Hollerith merged his young company with two othi firms to become the Computing-Tabulating-Recording (C T.R.) Com- pany C T R. (whichdpventually became the International Business Machines Corporation) had four basic units to offer.a keypunch for putting holes in cards, a hand-operated gang punch for coding repeti- rI tive data into several cards at the same time, a vertical sorter for arranging cards in selected groups, and altabulating machine forcom- piling the data punched into cards Over the next one-half century the roster of punched card 13roe- essing machines increased tremendously. The Remington Rand Cor- poratipn entered the field.and;along with IBM, offered devices for sorting, punching, verifying, merging, collating, reproducing, printing, tabulating, and calculating. All of theseachines, however, were dependent on electrical impulses to moveeghanical components

15 COMPUTERS AND SEMICONDUCTORS EARL ELECTRONIC COMPUTERS Althciugh electromechanical machines performedyeoman service for more than 40 years, theywere limited in both speed and flexibility. To move gears ane manipulate cards takestime; and because each machine was designed to perform onlya specific function, evenpore time was required to move the cards fromonevice to another to complete a given processing chore. Christopher Evans writes that , in the 1930s a shift was beginningto occur, and the threads of the problem were being gatheredtogether, quite independently, by a number of workersin various parts of the world. In the United States, large organizations such as IBM and Bell Telephonewere at work. In England the thrustwas coming from an indivi- , dual, the mathematician Alan Turing, whose paper "On Computable Numbers," publishedin 1936, sent a jolt of enlightenment among the cognoscenti:InGer- many, the threads were in the hands of ayoung I engineer named Konrad Zuse, who had madeup his mind not only to design a universalcomputer, but also to build one. . There.is some debate as to who first actuallycame up with the idea of an electronic computer. As far backas 1915 James Bryce, a consultant to C.T.R.,-had at least suggestedusing vacuum tubes for data processing purposes; and George Stibitz of BellTelephone Laboratories designed a computer using relay circuitry.But it appears that Zuse,an engineering student working on his doctoral thesisat theCniversity of Berlin in Charlottenburg,was slightly ahead of the others in the building of a working machine. In 1936 Zuse announced hewasgikiingup his job as a design engineer to build a computer. Rather than fabricatecomponents, he decided touse inexpensive, off-the-shelf parts. Although Zuse claims Konrad Zuse he was ncita,waie of Bablzage and his plaris for (1910- ) \ a Difference Engine, IBM Archives there are soe interesting parallels in design: both machines would have a memory, arithmetic section,'output, and be capable of being ' programmed for any kind of mathematicaljob. Zuse succeeded where Babbage faileOver the next few years the German' engineer built several successrulcalculatorseach one an improvement over the Previous machine. One of the ..... calculators, the 13, was used by a German aircraft manufacturerduring World War II to solve wing flutter problems; another, the Z4,-hadsome vacuum tubes to help it speed calculations foraircraft and missile design. Zuse

164. ONLY ONE SCIENCE

25 N

was apparently well on his wart° a full-scale computer when the wartime government decided to put its dwindliLgmoney and technical manpower into other areas.of research. Soratscientists today wonder what effect Zuse's computer would have hadon the war's outcome if he had been able to complete it. About the same time in the United States, Howard Aiken of Har- vard, With IBM's financial and technical help, built the electromechanical Mark I calculator which was essentigllya linking of 78 individual adding machines and tabulators. The Harvard calculator occupieda gymnasium-size room and sounded, in the words of writer-physicist Jeremy Bernstein, "like a roomful of ladies knitting." The clicking sounds came from the rapid opening and closing of its 3,300 switches. During 15 years ()fuse, the Mark I generateda huge amount of infor- matioin that was used, among other things, forinore accurate computa- n of the moon's prbit.

Howard Aiken The first true electronic digital computerthat is, a machine using (1900-1973) vacuum tubes for the generation and control of its electrical I B:M r. luu's impulseswas the Electionic Numerical Integrator and Computer, more commonly calleckENIAC4. ENIAC was developed at the Moore School of Engineering of the University of Pennsylvania by J. Presper Eckert, Jr. and JO)? D. Mauchly for the United States Army's Ord- The Nlark I calculator was built inance Departmetd It was a huge machine containing 18,800 vacuum between 1939 and 1944 (with finamial and technical help from tubes, and its inventors spent a good part of the first 21/2 years just IBM) by Howard Aiken of Har- soldering the 500,000 connections neededfor the tubes. ENIACcon- vard University sumed huge amounts of electrical power, and its glowing tubesgen- Smithsonian _I -

&&&&&& ......

17 COMPUTERS AND SEMICONDUCTORS

24 . erated a great deal of heat. It is said that.every time themachine . 1 was turned on, three of its tubes burned out, and the-lights inan area of Philadelphia dimmed momentarily. ..r

Spare plug-in units such asthis one were used to maintain the 18,800 vacuum tubes in the kIVJAC electronic digital corn- pilfera computer which required frequent maintenance 5,mthsolian

ENIAC, the first digital electronic computer, was invented by JPresper Eckert, Jr(left foreground) and John Maunchly (center foregrourid)at the Moore Schobl of Engineering of the Uni- versity of Pennsylvania in 1946, IBM Archives ,' In operation, ENIAC could perform 5,000 additionsp r second. All \ , internal functions were conducted by electrical impulses eneratedat 3 the rate of 100,000 per second. And while'this rate is barea crawl compared with the speed of present-day computers, itwas a tremen, dous advance over the capabilities of previous electromechanical machines. ENIAC's primaly job was to solve ballistics problems for the Aberdeen Proving Grounds in Maryland, where itsaw service from 1947 to 1955. During this period Eckert and Mauchly developed still another computer called BINACa loosely formedacronym for Binary Computerwhich eventually became the foreruriner of Rem-

...) 18 ONLY ONE SCIENCE

25 ington Rand Corporation's highly successful UNIVACUniversal Computerseries. Despiteits far-reaching features, though, ENIAC lacked the one elementIthat was needed to break Open the development of computers. It did not have a stored program memory!Operating instructions foi A the machine were recorded by the manual placement of plug wires. As

a result, once d'ata were entered into the machine, they had to progress , accortling to the paths laid down by these preset devices.

The Stored Program The criticalp4vance into stored program computers resulted Concept largely from the work of mathematicians involved in government proj- Acts to build machines intended primarilyjo solve military-related s'Cientific problems. In the early 1940s John von Neumann was one of the permanent members of the Institute for Advanced Study (IAS) in Princeton, New Jersey. Htolso,served as a consultant to the Army's Aberdeen Proving Grounds,as well as to the Atomic Energy Co ' mission's (AEC) Los Alamot Scientific Laboratory. t was his association with Aberdeen that led von Neumann a than e meeting with Herman Goldstine, then a lieutenant in th Unite States Army serving as liaison for the ENIAC team. In th sum- mer of 1944, the two men were waiting for'a train on the Aberdeen railroad platform. Galdstine, a mathematician himself, recognized von Neumann and introduced himself. Von Neumann asked the John Von Neumann LieutenarK what kind of work he was doing for the Army. Goldstine (1903-1957) replied be was helping to build an electronic computer that could per- IBM Archives ' form abdut 300 multiplications per second.

Von Neumann, as Goldstine later put it, was "galvanized." He . immediately saw the possibilitievf applying such a machine tocom- putational, problems in weapons design. "We got on the train together," \Goldstine recalls, "and from Aberdeen to Philadelphia he pumped me for details." Because of that chance encounter, von Neumann and his colleague at the IAS, Arthur W. Burks, entered into an active collaboration with Goldstine, Eckert, and Mauchly. Between 1946 and 1948, these menalong with several other scientists working at different institutionspublished papers on computer design and program planning: And from this work emerged the concept of the stored program, a landmark idea that was translated intoreality in a computer built at the IAS in 1952 under Von Neumann's direction. Several other machines of note, designed or built during the 1940s, might qualify as the first stored program digital computer. Among

19 COMPUTERS AND SEMICONDUCTORS

2 26 I I'

them was the National Bureau of Standards SEACbdilt by Sam Alex- ander for the .Inited States Army; the EDSACconstructed at Cam- bridge UniverTity in England; and the EDV ACat the Uhiversity of Pennsylvania. All had features that allowedat least some program sequences to be stored in tlie.same manner as datawere stored. ) a Regardless of which machinecame first, it was eviaent thatcom- puters could be extremely valuable tools for solution of sophisticated applied science problems. During the late1940s and the early 1950s, the AEC funded at least fivemore pioneering systems which became part of computer lore: the MANIAC at Los Alamos;the ILLIAC at the University of Illinois; the AVIDACat Argonne National Laboratory; the ORACLE at Oak Ridge; and,theJOHNNIACnamed affec- tionately for von Neumannat the RANDCorporation. During the .early 1950s these machines contributed immeasurablyto computer technology, and many of their individualinnovations found their way into the general purpose computers of IBM andRemingtbn Rand Cor- poration. "Perhaps our most important contribution,"observes David H. Jacobsohn, who workedon the AVIDAC at Argonne National Laboratory, "was that we convinced industrythere was an important scientific market for computers. At thetime, however,we had no ticeexcept to build these machines ourselves." With the ability to storeprctgrams in the computer, the pace of development accelerated tremendously. SeveralUnited States companies were now building computers f he general marketplace. In Great Britain, which for a fewyears led the wo d in computing sci- fiN ence, powerful data processinemachines appearin Cambridge, at Manchester University and 4. the National Physical Laboratory.A large English food corporation pioneered in theapplication of computer technology to commercial uses. i These early systems, however, had fro distinctdisadvantages: their enormous cost and the short life-span of their thbusandsof vacuungiabes. "You had to havea team of service engineers on hand at all times in order to keep up with burnedout tubes," recalls an early ''user. g r. . .

----*.

20 QNLY ONE SCIENCE - r SOLID-STATE TECHNOLOGY - / The vacuum tube, with its hair-thin wires sealed inside a fragile glass envelope, was a delicate and temperamental component at best. Even under optimum conditions,it had only alimited life-span. A large flow of electrical current was needed to get its filament 'tot enough to boil off electrons, but if this heat were not dissipated quickly and efficiently, the tube woulAburn out. Moreover, the whole unit was extremely vulnerable to damage or destruction from careless handling or external shocks.

Semiconductors The solution to the fragility of early computers came from a long series of experiments with solid-state materials. As far back as 1874,a CerNn physicist had reported a peculiar flow of electrical current in certain kinds of minerals. These minerals eventually became known as "semiconductors" because they conducted energy better than insula- tors, but not as well as conductors did. Despite their unusual properties, semiconductors had remained,

.11% for the mosIpart, just a laboratory curiosity. One exceptioncame dur, ing the early days of radio when the semiconductor material, carborundum, was used as a "cat's whisker" detectorso-called because the two thin leads attached to the crystal resembled a cat's whiskers. The carborundum functioned like a subway turnstile, allowing electrons to flow more ea =e direction than the other. In'thisway, the crystal was abl vert the oscillating electron signarthat came in through the antenna into a useable one-way flow of current. When vacuum tubes were developed, however, they proved to be much more effective at this conversion. Moreover, tubes could also boost or amplify current. Because of these advantages, most of the practical work with solids was discontinued. Nevertheless, many scientists were-still interested in thednusual properties of semiconductors. The advent of qualltum mechanics gave the first real understanding of how electrons in metalswere free to move and conduct electricity. And in 1931 physicist H. A. Wilson published a theory of how electrons and holes in semiconductors and insu- lators gave those materials their electrical qualities. These insights stimulated scientistsat Bell Telepl-;one. Labora- tories Incorporated to take a closer look at semiconchictors as possible modulators of electrical communication signals. Early in 1940, for exainple,ussell S. Ohl, a staff member working with the semicpn- ductor si con, called several colleague into his office to watch an

21 COMPUTERS AND SEMICONDUCTORS

2 LThs unusual experiment. Ohl showed them a small piece of black.silicon solid to which he had soldered two metal contacts. When light froma flashlight was shone on a nartw region near the middle,a photoelec- tromotive force of about 0.5 volts was developed. "I didn't believe what I saw," recant a staff member, Walter H. Brattain, "until Ohl gave me a piece to work with in my own laboratory." What Ohl and Brattain were working with was, in effect, the first o p-n (positive-to-negative-flow) junction transistor. With the advent of World War II, however, there was little time topursue the phenomenon any further. Although solid-state silicon detectorswere emkloyed during the war i dar devices (most of this work was done in Engrand and at the Ra ation Laboratory of the Massachusetts Insti- tute of Technology), th detectors could rectify only ceitain high- ,- frequency signals. VAuum tubes were still required'for amplification. In January of 1946 scientific research on semiconductorswas , resumed in earnest at Bell Laboratories, with the directed goal of finding a solid-state amplifier. Among those who participated in this land- taark work were William. Shockley, Walter Brattain, and John rdeen. Brattain wrote: At our first meeting, the group realized that in spite of all the work done before and during the war, we were still far from a real understanding. One reason was that copper oxide and other semiconductors on which early work had been done were very complicated-solids. Sili- con and germanium were the simplest, and the decision was to try to understand these first. Our work was directed toward a fundamental understanding of the problem though we were well aware of the technical importance of a semiconductor amplifier if one could be,; made. In December 1947, Shockley sent a casual note to a few colleagues inviting theni to observe "some effects" the research team hadcome across during an experiment with a device that contained gold contacts and a germanium semiconductor base. "I hope youcan break away and core e," he added. This may have been the understatemee of the age. What Shockley's team demonstrated a few days later was the "transistor effect" by which they could control the movement of electrons in a* semiconductor material through the influence of an outside electrical 4 field. The action of the device, which was named a "transistor,"wits explained this way: The transistor's amplification process can be understood in terms of the discovery that the input point is surrounded by an "area of interaction." Within this

Z2 ONLY ONE SCIENCE 2J area the electronic structure of the semiconductor is modified by.,the input current. Now, if the output point is placed in this area, the output current can be controlled by the input cement. This control of output current is the basic mechanism of amplification.

The first transistor wast.developed by William Shockley The,original transistor was called a point-contact transistor (seated), Walter Brattam (standing rIgh,t), and John Bardeen because it w0 essentially a wafer of germanium with two pointed- (standing left) at-Bell Labora- wire contacts loCated close together on one side. Shockley went on to toriesAlthough primitive by work out the theory ofn-p, n-p-n, today's standards, this early and p -n -p transistors and also to appar us demonstrated the tran- design the junction transistor which, in many ways, was more effec- sistorfect clearly enough to tive and efficient than the earlier types. In a very real sense, the generat riormous interest in further, research junction transistor sparked a technological revolution that has since BrILLabotaiories , changed the way people live. But demonstrating an effect and actually producing quantities of workable transistors were two very different things. Early researchers found that it Was almost impossible to predict how a given crystal would conduct current. Some allowed energy to move only in a positive direction; some permitted only negative flow; and some allowed two-way movement. Why the differences? Intensive research at several companies showed that crysrals of germanium or silicon taken from nature almost always have a small nmber of foreign atoms locked among their molecules. Because these foreign atoms have either more or fewer electrons in their orbits, they change the electrical characteristics of the crystal. Thus, if was difficult to tell in advance how a given transistor would behave in an electronic device.

COMPUTERS AND SEMICONDUCTORS

30 The people at Bell Laboratoriesmost particularly GorclOn Teal and Bmest Buehlerbegan to study this problem and soon discovered they could "grow" single crystals of silicon in the laboratory. In this manner they could.control the foreign atoms in the molecule and predict how the crystal would Conduct current. The key manufacturing process, however, turned out to be "zone refining," developed by Bell Laboratories' W. G. Pfann. As he explained it: The method of zone refining consists of slowly passing a series of molten zones through a relatively long ingot of impure solid. As a molten zone advances, impure solid melts at its leading interface, and purified solid freezes at its trailing interface. Each molten zone which _/ passes through the ingot carries a fraction of the impurity toward thepnd of the ingot. The purification increases with alb'number of zone passes. Germanium purified by zone refining is probably the purest known manufactured material.

W G Pfann (left), inventor of the so-called "zone -rufining" process, is shown here with his associate, J H Scuff, who is holding a large single crystal of germanium purified by this technique BeULaboratories

ONLY ONE SCIENCE In fact, crystals grown this way proved to beso pure that only 1 atom in 10 billion (equivalent to a pinch onalt in 38 carloads of sugar) was an impurity. The result was long, pencil :diameterloaves" of semiconductorusually germaniumwhichwere then sliced-If-Ito individual transitor pieces. To control the electrical characteristics, Bell Laboratories and Texas Instruments, among others, developed several methodsof "doping" the loaf with foreign atoms. Oneway to achieve the effect was to dope loaves with a few atoms of arsenic. Because arsenichas five electrons in its atomic configurationone more electron than is found in geimaniumthe fifth particle was free to "roamCand producen-type conduction Doping with gallium, which has only three electrons,left "holes" or "escape coutes2rthe molecularstructure which gave the crystal a p-type of electri al flow. Solid-state transistors could, therefore, be designed.to perform most of the functions formerly handled by vacuum tubes.

Integrated Circuits Although transistors were originally designed for telephone switching applications, they arrived at an evenmore opportune time for the computer industry. The extremely high coruponent failurerater, in computers wititthousands of vacuum tubesbecamea thing of the past When computer manufacturers started replacing tubes with transistors, t1.4 failure rate fell from once every few hours toonce every feW days at most. By the end of the 1950s, a second generation of computers had lappeared, using gerrhanium transistortheir basic active circuitele- titnent At the same time, storage mediums for computers were improving rapidlyfrom tubes, to mercury delay lines and magnetic drums, to magnetic core technology (developecLat MIT by JayForrester). In this latter techniquestill widely used in computersinformationis stored as magnetization in a tiny doughnut-shaped ferritecore As a , result, a large yolumeof instructionscan be stored in the computer, making the system faster, more dependable, andmore flexible. All of these technological improvements served to make thecorn- - puter a much more desirable tool. As Time Magazine reported, "No,one took to the (new) computer more eagerly or saw its usefulnessmore quickly than the businessman." General Electric became the firstcoin* mercial organization to acquire a data proCeFsing system; andmany other major firms, more particularly insurance companies with their huge information handling needs; followed withina few years. Nevertheless, it was the military that played an especially significant role in early computer development. The huge IBM STRETCHsystem,

25 COMPUTERS AND SEMICONDUCTORS cool

0 Doping a Silicon Chip for example, with 150,000 transistors capable of executing 100 billion instructions per day, was built in response to the Ballistic Missile Early Warning System's requirements for almost-instant data analysis and Pnototeisist computation. UNIVAC III and several large transistorized systems Oxide from Control Data also were developed for equally sophisticated scientific applications./ At first, manufactti ers looked for clever ways of packaging

Silicon transistorized circuits. One of the earliest featured transistors and Substrate P I printed circuit patterns wired together on cardstwhich could be a., plugged into, or taken out of, a main frame/Thus, if a failure did occur, and .'ated :oak photo-resist, an emulsion that ha, den, the faulty circuit could be replaced easily. / 14V),1 Lo,ltact lth ilzht But even occasional failures were much too frequent for many users, particularly when the computer was a keyelemeny1na defense 2 Unresiolet//// Light system. Analysis showed that most of the problems occurred ih theL. First Photornask interconnections rather than the transistors themselves. in the solder .,, ,,. joints, at the plug-and-socket connections, at wire-wrap joints, and so forth. And the next developmental step was to find a way of minimizing or eliminatikg these troublesome spots. Exposed s photoresist A new industry had developed in the UnUed States in the early photoresist 1950sthe semiconductor industry. In 1956 William Shockley, co- Ultra, 1,,lo hqht rs l,eamedar inventor of the transistor, left Bell Laboratories to form the Shockley . the wafer andtnolec's hun- ched:, of turd "(Ask trattrr r. Transistor Division of Beckman Instruments in Palo Alto, California. grit'hE rfac., the A year later, eight of Shockley's best engineers left, and with the back- ing of Fairchild Camera and Instrument Corporation started a division 3 which became known as Fairchild Semiconductor in the same area of 4 Unexposed phOtorillai Santa Clara County. Within a few years, so many new semiconductor removed corporations had been set up in and around Palo Alto that the region

became known as "Silicon Valley. One of the redeeming features of - semiconductor research, developrffent, and manufacture was that a lew firm could be startecl with relatively modest amounts of venture &pita'. Even as Fairchila4miconductor prospered and grew, several flitu_af,ris de, el,jed- of that company's best engi eers left to start their own operations. washed In a special solvent, reffil,1-1/1g the photoresist except In areas struck by the It did not take long for this activity to start paying off. In 1953, for ultra.r litfightA, litre tht instance, a Radio Corporation of America physicist applied for a patent phott, rosist [las bee', roilow,i on a circuit that could be fabricated on a single block of germanium, a the oxuie is now exposed year or two later several English scientists extended the concept; in 1958 an engirieer -at Texas Instruments succeeded in making an integrated circuit (IC); and a few months after that a team at Fairchild Semiconductor began work on what today are called microelectronic circuits. Simply put, an integrated circuit is one in which all the active and

26 ONLY ONE SCIENCE

33 passive'elements are formed together on a small chip of semiconductor material. There are no separately installed joints, wires, or other parts that can 'break or become unraveled. Nevertheless,even an integrated circuit can failusually as a result of a manufacturing defect. The answer to this problem came from IBM engineers who deSigned a high-speed machine capable of testing the tiny chips before theyare installed in a computer. Thus, once an integrated circuit is insidea data ett;11.14 hat"' ronc-,,, the processing system, its failure rate is virtually nil. axial' tho,eel(r,25eaar,aa re: ea11.1 ,aNtrial,Ilar Although the early ICs were formed on a germanium base, itsoon became evident that silicon offeredsome distinct advantages The reason lie's in the basic process. Most integrated circuits today are made by the same kind of photoetching techniques used in metalworking Phosphorus arsenic or antimony industries. To start, engineers coat the surface ofa semiconductor diffusion wafer with a thin photosensitive film called "photoresist." What this does, in effect, is to turn the wafer into a kind of photo contactpaper. Then a mask or stencil of the circuit pattern that 'has been photograph- Electrical ically reduced from a large drawing is placed characlarirlict over the surface and changed In.) exposed to ultraviolet light The light has the effect of either hardening

Pi or softening partiof the material. The softer material is washed away rth ,1".,myrt7 Implomt, puttoN tt into a dr"usum by.a solvent, leaving the semiconductor surface open or exposed in the "Loma" al pre 1,1 desired pattern. As a final step, doping impurities or metal for the 1,t1r14,f ape, aria /11rsirix: interconnection lines are deposited on the exposed areas. I' at a temperature c4 rri. re 441,1 .1,1 The photosensitive material, however, is insufficient by itself to ;ICIIr h IS 1.,,,Ce

Large-Scale Integration Initially, single integrated circuits were replicated scores of times over on the starting wafer, and n the wafer`was diced into individual chips. It soon became ev dent, though, that a much more func-

27 COMPUTERS AND SEMICONDUCTORS /- tional device could be formed by interconnecting several of these circuits. The result was what industry people called Large-Scale Integra- tion (LSI), in which a large number of circuits, each witha different futiction, are formed together on a single, Va -inch-square silicon chip. La's, ppearedin the late 1960s, gave thecomputer industry a huge boost. Because the first two generations of data processing machines had tremendous cost, heavy power needs, and huge heat dissipation problems, the only organizations that could even consider installing them were large government agencies, major scientific centers, and leadi9g corporations. With LSI technology, however, the size, cost, electrical drain, and heat generation of computers were reduced to the point where thesys- tems came within reach of most medium-sized andsome smaller companies Not only did this mean a vastly expanded Marketplace for computer manufacturers, but it also opened the gates for hundreds of entirely new firms Among them were consulting organizationsto help a company plan its data processing needs and install the equipment; "software" houses to write the special programs required by industry; peripheral equipment manufacturers o produce data storage facilities, terminals, and other "plug-compatibl devices which could be linked to a main frame, and a still growing nu er of smaller manufacturers making special function and miniaturized omputers. Thus, what started as technology dominated by justa few firms soon involved hundreds of different companies.

Microprocessors When Bell Laboratories announced the transistor back in 1948,no one could really foresee the revolution in electronics that would follow. Yet, by 1960, even the tiny transistor seemed bulky whencom- pared with the integrated circuits and the large scale integration which crowded increasing numbers of components onto an ever-shrinking silicon square. Still another quantum jump in the technology of minia- turization can be witnessed todaythe emergence of the so-called "miracle chip" on which both logic and memory circuitryare contained There is some question as to whether Texas Instuments or The Intel Corporation was first in developing microprocessors The most widely repeated story, hOwever, concerns a project which took place at Intel back in 1969. As Time,Magazine described the events: Fresh out of Stanford University, where he had been a research associate, M E. "Ted" Hoff was placed [by Intel] in charge of producing a set of miniature components for programmable desk-top calculators that a

~~28 ONLY ONE SCIENCE~~

35 Japanese firm- planried to market. After studying the circuitry proposed by the Japanese designers, the shy, self- effacing Hoff knew that he had a problem. As he recalls: "The Calculators required a large number of chips, all of them quite expensive, and it looked, quite frankly, as if it would tax all our design iapability." To solve the problem, Hoffcame up with a novel idea. Why not place most of the calculator's arithmetic and logiccircuitry on one tiny chip of silicon? After wrestling with the design, Hoff and hisassociates at Intel were able to place nearly all the elements ofa central processing unit (CPU)the "main frame," as it is called in large scalesystems on a single chip. Finally unveiled in 1971, the one-chip CWJ,now called a microprocessor, contained 2,250 transistors inan area barely 1/6 of an inch long and 1/8 of an inch wide. In computationalpower, the tiny microprocessor vas almost as powerfulas the original ENIAC, and perforthed 4s well as many IBM machines of the1960s. It was, as the cOmpany was quick to advertise, "anew era of integrated elearonics...a micro-programmablecomputer on a chip." To most people, the word "larger" usuallymeans "faster" or "more powerful." A large, eight-cylinder engine,, for instance,produces more horsepower than a small, four- cylinder one. A mammoth steam shovel can probably dig out more dirt in one scoop thana person with a spade could move in many months. In the familiarmechanical world, a larger machine almost always doesa job faster, more easily, and more econorhically than a smallerone. In microelectronics, however, thereverse holds true. Chips are hundreds or thousands of times smaller than thevacuum tubes and transistors they have replaced. Yet everything about them is faster, more powerful, more reliable, and more economical. A desk-sizecom- puter today can often produce the same amount of workas a computer , that once occupied an entire room. In computerg, becausea major hill- iting factor in the rate of computing is the timeVequiredto move electric signals within the machine, smallis beautiful. Asmore components are packed onto a chip, signals travel shorter distances and calculating speeds go up; and as density and seed increase, computing costs go down. Theresult: many times faster processing ata fraction of the cost. For those applications where small size is in itselfa desirable end,it the chip is made to order. Electronic Wrist watched, versatileRocketcal- culators, portable foreign language translators computer- controlled microwave ovens, automatically focused cameras,,palm-size color television sets, automobile trip computers, engine monitoring computersall of the new miniaturized computational deviceson the,

~~29 COMPUTERS AND SEMICONDUCTORS~~

~~36 market todayowe their existence to the thousands of transistors, memory cells, and passive elements crammed onto the surface of a silicon chip small enough to passithrough the eye of a needle.~~

PROGRESS-1AI PROGRAMMING As circuitry got smaller andore reliable, an equally important change took place in the programmg of computers. When data processing systems first appeared, most scientists neededan intermediary trained in programming to use them. Even with this kind of help,it took months or years to write a major program. Thereason was that even the simplest problems involved the laborious and error-prone process of setting down long segments of instruions in the precise order needed by the machine. As a result, mos organizations hadto hire largeand expensiveprogramming sffs if they wanted to get their system "on the air" in a reasonable amont of time.. The first major improvement came from the work of Grace Hopper, who started programming with Howard Aiken's MARK I computer group. She later served as a senior mathematician for the short-lived Eckert-Mauchly Computer Corporation and finally did some of her most productive work for the Remington Rand UNIVAC group At Remington Rand, she developed the first practical "com- piler" program which translated relatively compact instructions, such as "ADD C," into detailed binary code. More than anything else, her work showed that the computer itself could be used to do itsown program writing, translating more or less standard English terms into binary numbers.

Grace Murray Hopper The next major advance came with thedevelopment of so-called sT190(7,, ) problem-oriented languages. The first of these, still widely used,was IBM Arch:yes FORTRAN, created by John Backus and his IBM colleagues in 1956. FORTRAN gave scientists, in particular, the ability to write theirown programs in algebra-like expressions. Other math-oriented languages, such as ALGOL, PL/1, ancloAPL, followed and had a profound influence on scientific computing. With ALGOL, for example, a complex program could be written in a few instructions instead of in several FORTRAN pages. Thus, programming became faster and "debug: ging" easier, witlf the net result of a drastic reduction in turn-around time between conceptualization of a problem and testing of thecon- cept on a computer. Today many of theinstructions which formerly had to be written by a programmer are preprogrammed intoa computer by its manufac-

~~30 ONLY ONE SCIENCE~~

~~1. 37 v~~

turer. Nevertheless, the art and science of getting a machine to do what is wanted still requires a skilled individual working with a pirticular programming ldng-iiage. Every computing system, from the largest distributed processing networks tettrtKpact stand-alone computers,consists of five elements: input, storage, processing, output, and control. To make sure a job is performedproperly, thelHogrammer must guide the system through all of these areas in aprecise step-by-stepmanner. While programming still requires a considerable amount of human skill and patience, an increasing amount of this work isnow perfotmed by the computer itself: Among the developments that have made this possible are the following: simplified programming languages and translators which take English-like statements and translate them directly into 0's and fs; program libraries from whicha user can select prewritten routines and piece them together to form all or part of a complex complete program; structured programming sys- The one -chip computer was de- velopedfor a variety of tele- tems Which allow an untrained individual to develop a'program by t. OMP711.0111_ations appfiLations answering machine-generated questions, preprogrammed memory Here the size of the chip is compared- with that of an ordinary chips which can "sense" certain internal activities and then supply the paper clip needed data management routines; and error-correcting codes, Bell Laboratorws developed by Richard Hamming.

TRENDSiN COMPUTER USE In the nineteenth century, Charles Babbage speculated that science would eventually "grind to a halt" because of a lack of calculating power. Instead, the opposite has happened; the pace of scientific 'development has picked up to the point where peopleare now being inundated with new information. This is true not only in science and technology but also in virtually every area of business and industry; Thus, ohe critical need today is to find ways of managing and using the ever - increasing information flow. Clearly, computersare tools'almosi ideally designed for this kind of work. With their speed, storage capacity, display ability, andease of programmIng, data processing system'sare now finding their way into application areas not even dreamed of a decade ago. Here, for example, are just a few applications to illustrate some of the directions in which computers are now moving: ,.. ( Scientific Information Computers are being used to store nd to retrieve the huge Storage and Retrieval amounts of scientific information on h nd today. (It has been

~~31 COMPUTERS AND SEMI ONDUCTORS -40 1~~

.,(r nr1 J O I ,estimated that the amount of this information has roughly doubled during every decade of this cent ry',) A major application in the scientific and technological fields as een the development of information storage and retrieval systems. Generally speaking, the information storage and retrieval systems being developed today feature huge data baseswith a given bpse usually confined toa specific scientific disciplinewhich cal4 be reached through a variety of remote terminals. In most cases, the user needs very little trainingor computer knowledge to get at the material. "All it takes isa few key statements punched out on a typewriter keyboard," explains acancer, researcher at fire University of California, Los Angeles, School of Medicine, "and the request goes directly to the MEDLARS (Medical Library and Retrieval System) computer in Washington, D.C." After that, it is only a matter of seconds before a title, abstract, or whole paperdepending on what is requestedis flashed back via the same channels, and : either displayed or printed out by the same terminal.

111,00 Analysis of Satellite- Computers are being used with a program designed to make better Gathered Data uses of the earth's resources. The program also is used to correct map- ping information collected by satellite. As the National Aeronautics and Space Administration's LANDSAT satellite circles the globe each day, its cameras and other sensor devices map sesitiops of the earth's surface and transmit digitized information to a land station. Ordi- narily, such things as satellite roll, pitch and yaw, earth rotation, and sensor errors would make these digitized pictures very difficult to read. But the corrective "lens" Oflhe computer's program is able to reconsti- tute each of its 115-mile-square pictures with remarkable clarity even filling in sections that are missed by the cameras. Under certain conditions, even a single vehicle can be identified. One key use for this satellite-collected computer-processed data is in agriculture. Through the progranl* the computer can help to identify acreage under cultivation by type of crop, can be used in the assess- 4, ment of damage from floods or insect infestation, and, most significantly, can be used to predict crop yields. In 1979, for example, satellite data were used to estimate wheat production in the United States and Canadaa figure which turned out to be more than 95 percent correct when the actual crop was harvested. By the end of 1980 the same program was applied on a global scale to make predictions about __..w.orld-output-of-lAtheat-c-ramrcottonrricerand soybeans. Another important area for such systems, alrea' dy being used to some extent by a few oil companies, is in cutting the time and effort

32 ONLY ONE SCIENCE 3J With the aid of LANDSAT im- agery, a United States Depart- ment of Agriculture statistician is developing new mathematical procedures to predict crop yields USDA involved in energy exploration. Using computer-produced picturesof seismic reflection patterns, geologists can spot formations in theearth which may contain oil or as deposits. With this kind ofinformation, they can theri conntrate their land exploration efforts to increase chances of a find. The importance of thecoputer in analyzing data to help locate oil a d gas 'deposits is describen the chapter, "Seismic Expla1Tron for Oil and Gas.") Other areas for satellite-computei applications include:land mai,- ping, coastal zone management, monitoring waterrun -off from mountain snows, tracking icebergs, urban planning, selectingright-of-way corridors for pipelines, power lines, and highways, and gatheringintel- ligence inforatian. In the near future, satellites in stationary orbits over the oceans of the world will, through land-based computers, help to keep track of ship traffic.

Industrial Applications In industry, goods-in-transit systemsare already proving to be a rich vein for computers. Presently, many companiesuse computers to assist in the complex process of rating freight bills, improving utiliza- tion of trucks or railroad cars, and scheduling preventive maintenance for their fleets. The greatest potentiabfor savings, however, lies in scheduling goods directly from shipper to consignee rather than simply from yard to yard. Studies show that a leaded rail carmoves toward its destina- tion only about 11 percent of the time, leavingroom for a tremendous

~~33 COMPUTERS AND SEMICONDUCTORS~~

6 . amount of improvement. A mere 1 percent improvement, for example, 'could free up,to 17,000 rail cars. Thus, computer applications which help to increase resource productivity are among the most important in industry today. Another indUstrial application now clearly coming into its own is computer-aided design (CAD), sometimes referred to as engineering

graphics, Since CAD came into existence about 10 years ago, most of . the large automotive and aerospace companies Wave made it p fine art. What is significant today, however, is that many intermediate-sized manufacturers are now reaping product improvements and cost savings through CAD. One CAD application being used at a heavy equipment manufacturing firm, for instance, has eliminated up to 50 steps between the creafibe of a basic'asic drawing and the actual cutting of a part on a machine tool. And every one of these saved steps can be translated into dollar terms. Computers can even be used to design themselves. As more complexity has been needed in very small integrated circuits, CAD has been used successfully to evalCate the optimal arrangement of subcircuits and their connections and interconnections.

~~Computer-aided design (CAD) has been of great service to the engineering-desigtrindustry by reducing the costs of production through the elimmattorrof draft- ing steps pm Archives~~

~~34 ONLY ONE SCIENCE~~

~~,41~~

"7 Commercial Although business has lOng used computers for accounting and Applications recordkeeping functions, the accelerating trend today is toward electronic, "paperless" offices. Many firms now use computers for text 111 entry, informati retrieval, composition, and for electronic ..,, i.e. transmission of letteimages, often via satellite, from computer to computer within a copany. In addition, many large legal organizations now use comp terized systems for both storage and retrieval of N.0 pertinent legal da a. Point-of-sale computer applications, such as the terminals used in retail and grocery fizins for credit card transactions, check caking authorization, debit card activities, and automatic laser-beam 'scanning li. of bar codes, are gradually replacing the traditional cash register. Such C-systems provide productivity improvements, timely management information, and better control over store operations in one recording A step. Yet the trend is not limited to these areas alone. Drug stores are now installing pharmacy systems to improve the process of filling prescriptions. When a customer purchases a prescription drug, the A computer provides information such as side-effects, expiration dates, prices, and a receipted record for the customer's tax purposes.

. Medical Applications Hospitals and physicians are finding more and more uses for computers. Computerized Tomography (CT) scanners now use Computers to intensify and analyze cross-sectional and longitudinal X-ray pictures of the body. (CT scanners are discussed in more detail in the chapter, "X Rays for Medical Diagnosis.") Computersalso monitor

-. and signal events in the Intensive Care Unit (ICI)), help to provide rapid and accurate readout of laboratory tests,and keep track of patients' progress from the time they enter the hospital until long after .% they have been discharged. /-' L

Scientific Laboratory Today there is hardly a scientific laboratory that does not make Applications use of computers. Minicomputers, and even microcomputers, are o en integfal parts of such sophisticated experimental machines as .., a , NLear Magnetic Resonance (NMR) and Electric Paramagnetic Reso- nance (EPR).apparatuses. In the most advanced instrumentations, data are handled from their source to their ultimate disposition entirely by a of computer. The human researcher intervenes to direct the course of an experiment, interprresults, and generally manage the laboratory system. But all of pes ctions'are enhanced to some degree by a computer.

~~35 a.COMPUTERS Pap SEMICONDUCTORS V~~

Of r Finally, the computer is increasingly being used in the home. Originally, applications of computers away from the work placewere limited primarily to a handful of engineers, scientists, students, or sophisticated people in business who were obliged to do a lot of work at home.'With the advent of the microprocessor/ however, all sorts of 'convenience applicationsfrom home recipe planning toenergy savings systemsare becoming apparent. QUALITY OF LIFE Despite the vital nature of these information management functions and the usefulness of the devices made possible'by microprocessors, the very mention of computers and chips often elicits only a negative response. Some people blame computers for many of the more unpleasant aspects of modern life: billing errors, junk mail, commercial crime, job dislocation, and threats to privacY and inviduality. These problems should be viewed from the perspective of ate ology that is still growing, if not exploding. Although the victims o computer technology gone awry are often told, "The computer did it is actually the people who program and run computers who create most of the problems for which their machines are blamed. "Computers do only what they are told," says a data processing scientist, "and not an iota more." In the final analysis, it should be recognized that technology creates new products and services and, therefore, new job markets. Already, microprocessors have opened a home computer market; they have been responsible for scores of new calculator-type games and manrilifferent electronic children's toys; they have made possible many nep-miniaturized communications devices; and they are the ,keysto0 of a still-growing digital watch industry. "Nevertheless," says Bell Laboratories microprocessor chief Lee Thomas, "the most exciting applications Won't come until the kids who are still in high school and have grown up with pocket calculators and home com- putersbecome theengineers of the 1980s and 1990s." Thomas also 4feels that if American business fails to exploit this capability, a potentially enormous market could slip through its fingers.

THE FUTURE Despite the dramatic developments that have taken place in recent years, the world is still in an early stage of microelectronic technology. And from thperspective of more than 30 years of progress, a few conclusions stand out.

~~36 ONLY ONE SCIENCE~~

43 First, most of the devices and components used in computers today did not result from accidental discovery, but from intensive, mission-oriented research programs: In this field, typically, scientists, have deliberately set out to find alternatives to components already in use. Secondly, the unparalleled advances in semiconductor and com- putertechnology during this period were the products of interdisci- plinary teamwork. The complexities of integrated circuitsnot to mention the complexities of putting a computer togetherrequired the correlated effortseof'specialists,in materials, components, circuits, and systems. And the job of moving the circuit from a drawing board to its place in a computer requiretill another team of highly trained people. Finally, applying computers to an ever-expanding range of jobs is dependent on people whose skills were not even known 30 years ago. These peopleamong them programmers, systems analysts, and even

, a pew breed of sales representativeshave all contributed to the Suc- cess of this new, yet still growing industry. Few technologies in histgry have come so far so fast. Thirty years ago people were looking forward to file ",nuclear age." Howeverthe . computer's spectacular growth 7irtnurnbers, in power, and capability, in the variety of things it can do, and in the economy it provides when doing these things-r--has brought about the "computer age." What is more, it is a safe bet that this new era will be around for a long time to

~~, 37 COMPUTERS AND SEMICONDUCTORS /~~

~~.0.~~

~~4e,~~

* ,12 ASeismic Exploration

~~oJ 0' for Oil and Gas AV AVM~~

In a Connecticut Veterans Administration hospital, a man who has lost his eyesight listens carefully to stereo recordings of someone walking through hallways, in and out of rooms, and along city streets. His instructor points out important clues in the quality and loudness of the sounds. With this help, the sightless p-atient will be able to orient himself; he will soon be able to distinguish differences in the reflected sound of his footsteps or of his cane taps. He will be able to tell if he is walking along a corridor by the speed with which the echo of his footsteps returns from the nearby walls; he will learn to distinguish differences by the change in sound quality when he walks into an intersection of corridors; he willliahato listen to those echoes for clues to the size, the shape, and even the furnisning,s of, a room. In other words, the blind patient is being trained to sense and process information about his surroundings based on his experience of the reflection time of sound waves through air. Someone standing on the surface df the ground yho needs to learn about the structure of the earth 100 feet, 1,000 Feet, or even 1,000 miles below is like the blind man. Althougha sightless person could conceivably walk around the room touching walls and feeling sulTaces, a similar technique is clearly impractical for a geologist or geophysicist trying to learn about the structure of scores of cubic miles of earth. By setting ft expi,swns sitL One of the most useful methods of the modern geophysicist is analo- as the one shown on the oppo- sae page, geophysicists can gain gous to the technique that the patient at he Veterans Administration information ahviet the LorriposJ hospital was learning. The geophysicist, too, can learn about the world hot;of the various strata below the surface of the earth Such by interpreting the behavior of walesin.this case, seismic waves as Information helps them locate they travel through the earth. areas where oil and gas nay have accumulated The technology that allows the'geophisicist of today to see below (...)S.H H , s,141, the surface of the earth is based oh contributions of scientists from a

~~39 SEISMIC EXPLORATION FCIIIL AND GAS~~

~~4i .,~~

broad variety of scientific fields. Many of these scientists, suchas Leonardo da Vinci, Isaac Newton, and Christian Huygens,were trying to satisfy their curiosity about how things work, how the earth's surL face came to be, or how natural phenomena, especially light and sound, behave. Da Vinci's geological observations, Newton'scogception of the physical universe, and Huygens' unravelling ofsome of the

mysteries of the behavior of light, were combined hurldreds ofyears .. later by geophysicists trying to develop versatileand reliable methods of locating oil and gas deposits. In the work of A. many of the giants of science, others saw techniques and concepts that could beshaped and adapted to search for oil andgas and even to explore the inner structure lof the earth to its core. BASIC PRINCIPLES OF I SEISMIC . (--74 EXPLORATION Anyone who has felt the ground tremble underfoot has sensed clearly that vibrations travel through the earth. Whether itis a heavy truck, a subway train, an explosion, the demolition ofa building, or an earthquake that causes the ground to tremble, they all haveone thing in commonthey send waves (vibrations) traveling through the earth. More than 400 years ago, English soldiers besiegedat Exeter recognized and took advantage of this phenomenon. Concerned that theirenemies might breach their defenses by tunneling under the walls, the soldieis set out pans of water and watched them carefully for surface ripples caused by the impact of shovels digging below. The modernseismic methbd of exploration for gas and oil takes advantage of thesame principles of detecting vibrations, but modern technologicalprogress has produced vibration-detecting equipment faimore sophisticated than pans of water set out on the ground. Today, application of the seismic method to look for oilor gas uses explosives or other energy sources,scores of specially designed geophones, magnetic tape recorders capable of recordingup to 1,024 channels of sound simultaneously, and computersas powerful as those used in the space program. While itwas important for those soldiers at Exeter to knOw if their enemies were tunneling, they neededto know . only whether or not there was vibration. Interpretationwas simpleif there were vibrations that could not be explathed otherwise,the enemy was at work below. Today's geophysicist is looking much farther and deeper than those sixteenthLcentury soldiers. To reach useful conclu-.. sions,modern geophysicists use far more precise data. Still, the

~~40 ONLY ONE SCIENCE~~

~~47 FX/A-1;f-t,tivAliniffia~~

~~Recordings made from modern seismographs enhance the ability of geophysicists to make inter- pretations Mobil Oil Corporation, Merathon Oil Company e5.0 0 40 'Ne e -ei e a~~

~~4e .Sd~~

~~_ -C.~~

41 SEISMIC EXPLORATION FOR OIL AND GAS interpretation of the stacks of computerprintouts and graphs is simply that.. -an interpretation. Seismic explorationdoes not indicate where oil is, only where it is likely to be. Andin the case of seismic prospecting for oil, a decision to drillcart mean millions of wasted dollarsor it can mean the discovery of significant deposits of oil,the development of natural resources, andcorpprate prpfitsan awesome gamble to base on what is-nomore tharran'e`due.ited guess." For theguess to be as educated as possible, the geophysicist needsevery scrap of information and every clue available.

Formation and The ways that petroleum formedand accumulated have dictated Accymulation of the kinds of clues geophysicistsseek with their sophisticated equip- Petroleum ment Most geologists today believethat petroleum is formed from the remains of plants and animals whichlived millions ofyears ago in or along the coasts Of the shallowseas, swamps, and lakes that covered many parts of the earth. When these organisms died, theirremains eventually sank to the bottom andwere covered by sediments, such,as silt, shale, or limestone. As thesediments and organic matterwere buried deeper anadeeper, theywere subjected to increased temperature and pressure. The-sediments gra,tally formedrock, while bacteria, heat, pressure, and diverse naturalprocesses, which are still not totally understood, transformedthe organic material into oil andgas. The sedimentary rock in whic5the oil and gas were formed, commonly mudstone or shale, is called "source rock." Aslayers of sediments continued to accumulate under thewater, the newly formed oil andgas were subjected to increasing pressure. Thepressure caused the oil and gas to move into nearby porous sedimentary layers called"reservoir 'rocks" which were under lesspressure. Within theiRtf formation, oil and gas may also move upward becausethey are less dense than the water which fills the spaces in the rocks. The reservoir rocks, no matter whattheir compositionsand, sandstone, or limestonehave twovery impor6nt properties: they are porous and permeable. The oil moves into thepore spaces within these rocks in much the sameway that Water poured into a glass filled with sand finds its way into the emptyspaces befween the grains. The more porous the reservoir rock, the more oil and naturalgas it can hold. In addition to thesource rock where the oil and gas are formed and the reservoir rock in which theycan be stored, a third condition is necessary for significant oil or gas accumulationthepresence of a trap What happens when oil orgas is not trapped can be demonstrated

~~42 ONLY ONE SCIENCE~~

43 with a simple oil lamp. When a wick is put i,t1 the oil reservoir of the lamp, the oil saturates the wick. The longer the wick, the longer it takes for the oil to reach the end. Eventually, though, the wick will draw all of the oil from the lamp, resulting in a damp wick and a dry reservoir. With crude oil in the earth, the porous rock is Botht'h the reservoir and the wick. The oil, unless blocked, migrates upward and laterally through the porous rock layers until it is dispersed so thinly that, for all intents and purposes, it disappears. It may even reach the surface where the lighter fraction may escape into the atmosphere, as in the well-known case of the La Brea, tar pits. Luckily, oilcan be trapped in the reservoir rocks if the adjacent rock is nonporous. But How can this happen if the sedimentary layers are horizontal?

Almost 500 years ago Leonardo-GIA Vinci glimpsed part of the \ answer. The fossil seashells that he found in the Alps convinced him that the surface of the earth had undergone gigantic changes. During the long periods of time that it took for the formation of oil, segments -of the earth itself shifted. Huge areas of therplanet's surface rose and fell. Nev bodies of water appeared, and the bottoms of ancient bodies of water became high plains, deserts, and sometimes even mountains. Duringfthis time, the earth's crust buckled, -...racked, and folded, and faults developed. This geologic activity millions of years ago set the stage for modern technological society by forming traps for the accu- mulationof oil and natural gas. One common example of such a trapthe salt ome. The oil and gas around a salt dome were formed in an organica y rich sedimentary layer. BuckliN in the earth's crust, probably cause by salt pushing upward, deformed the original horizontal layering of the sedimentary material. This action of the salt dome created fractures in the distorted layers, trapping oil and gas around the Ome. The gas, less dense than oil, flowed to the top of-the trap. Today, the geophysicist who locates a salt dome knows that this is the kind of structure that might be associated with oil. Only a successful well will tell for sure. Breaks in the earth's crust, or faults, are another common form of hydrocarbon-trapping structure. Although the,break can be in any direction, the one commonly associated with,oil entrapments forms when the earth shifts vertically and causes a permeable layer to lie next to an impermeable layer. Again, the story begins with the deposition of organically rich sedimentary layers. The organic material begins to be transformed into oil and natural gas. As the layers shift in their ;elation to each other_ a fracture with displacementa faultmay be formed.

~~43 SEISMIC EXPLORATION FOR OIL AND GAS PRINCIPAL TYPES OF OIL TRAPS oil~~

~~(a) Fault , it- I --, $' 4 k.# I r--- --I!' shale crystalline limestonesandstone Typical petroleum traps (a) Fault (6) Salt dome. Impermeable rocks rmeable rocks~~

~~salt oil water~~

~~(b) Salt dome~~

~~44 ONLY ONE SCIENCE~~

5 As petroleum is formed, it tends to move generally upward along permeable beds. In a particular circumstance, for exami?le, perhaps the oil flow can move only toward the right and up until it reaches the fault itself.Because the right side of the fault consists of an impermeable layer, the oil and as must stop. A new reservoir is formed. There is an almost infinite variety of traps that might contain oil. What is im &ortant to remember is that while accumulations of are found only in traps, not X11 traps contain oil. The fact thata geologic structure could trap oil does not mean atilt it did. The proof is in the drilling.

Behavior of Seismic Before any kind of interpretation of information gained from Waves seismic waves is possible, how,ever, geophysicists need to understand the behavio,f Of these waves as they travel through the earth. Today it is known that as a seismic wave travels through the earth, a part of it is-

~~ENERGY GROUND RECORDING SOURCE SENSORS SYSTEM Truck-mounted Geophone Amplifiers &- hydraulic vibrator array tape recorder~~

~~Seismic reflection profiling using the Vibrosels® technique 7~~

~~45 SEISMIC EXPLORATION FOR OIL AND GAS~~

ti bounced back each time it encountersan interface between differenk layers of rock. This is a little like what happens whenripples set Cif by a pebble in a pond strike an object in the water. Someone looking at the water closely will see smaller ripples that have bounced off thatobject moving away frcim it. In the case ofa seismic exploration, the waves reflected from rock layers bounce upward and travelback to the surface where sensitive detectors, calldd geophones,record them. The geophone is similar to a microphone. It "hears"the vibrations in the ground and transmits them toa tape recorder. In the same way that the captain ofa ship learns the depth of the water by learning how long it takesa "ping" or sound wave to reach the ocean bottom and return, geophysicistsmeasure the time it takes the seismic wave from a small explosion to bounceback to their detectors. But, in dealing with a single and faytly uniform element water--the ship's captain hasa comparatively easy task. Sourtd travels through water with a known speed. Distance, then,is easy to calculate. Once the speed of sound in water is knownand the time it takes'for the echo to return is measured, distancecan be calculated: Within the earth, however, the application of thesame principles is far more complicated for a number ofreasons. In 1845 an Irish engineer, Robert Mallet, triedto discover the speed that sound travels through the earth.He got the wrong answer, but his experiment was so ingenious that hismethodology made the largest single contribution to the seismic methodin the nineteenth century. He buried a charge of gunpovider connectedto an electric detonator and placed bowls ofmercury various distances from the -sal.explosion. A spotlight on themercury illuminated the liquid so that with a telescope he couldsee ripples on the surface when the seismic wave reached the bowl and made it quiver. To time thewave he used a clock which was stirIrd electrically and simultaneouslywith the explosion. He could measure the distanceon the ground betwec4i the site of explosion (today called the shotpoint) and themercury (corresponding to today's electronic sensors). Bymeasuring the time it took for the seismicwave to reach the bowl; he could determine the speed of the wave. Mallet also made thevery important obs'ervation that different velocities were derived from differentsurface geologic materials. The observation that the velocity of theseismic wave was dependent on the sort of rock thrgh which the wave was traveling was to be of immense importance to twentieth-century geophysicists. For example, it was discovered that seismicwaves traveled through rock much faster than through soil. Because ofthe comparative insen- sitivity of the mercury, the actual velocities Mallet calculatedturped

~~46 ONLY ONE SCIENCE~~

5ti out to be too low. Nevertheless, his work was a success. In the middle of the nineteenth century Robert Mallet had,devised all of the basic elements of the seismic method. He had also discovered that the speed of a seismic wave was related to the material through which it was , traveling. Modern developments with their more sensitive and complex electronic components are, in a real sensft, refinements of Mallet's pioneering work. PROBLEMS ENCOUNTERED WHEN USING THE SEISMIC METHOD From the beginning, the problems of "seeing" underneath the surface of the earth confused the data. Mallet did not know that his mercury detectors were giving him inexact information because the mercury was not sensitive enough to signal the arrival of the first weak vibrations arriving at the bowls. Today, with all the advances in physics, geology, and electronics, geophysicists still have to face an impressive array of problems that grow out of the behavior of seismic waves. / Velocity The first problem facing the geophysicistis that, unlike water, the earth is made up of a variety of materials. In water, sound travels at approximately 5,000 feet per second, varying only slightly if the water is salt or fresh. But in the earth, sound travels anywhere from 1,500 ''' feet per second in some unconsolidated muds, clays, and sandy materials, to as fast as 25,000 feet per second in some dense, deeply buried rocks. Geophysicists refer to the speed of sound through a material as the "seismic velocity" of that substance. Seismic velocity is determined by the solid state crystl structure, the density, the amount of * Ore space, and the fluid or fl ids filling the pore spaces of the material. ilk In a sense, the seismic velocity of a rock layer is a signature identi- ying th,e-type of rock. That, in turn, can be an indicat r as to the,likeli- h hoodof the existence of the proper conditions for the resence of 'oil Seneath the surface.

* c. Refraction A second problem is also related to speed. Under most conditions, when a wave, including a seismic Wave, changes speed, it also cliahges 71direction; it bends. Physicists refer to this as a change in the direction of propagation or "refraction." Light waves behave the same way;

~~47 SEISMIC EXPLORATION FOR OIL AND GAS e~~

Alena ruler is put into water at an angle to the surface of the water, a the ruler appears to be bent at the point where it enters. As the light wave's travel from one medium, water, to another medium, air, they C- change speed; they are refracted if the path of the lightwave is at angle to the interface. The ruler appears bent7Since the layer of the earth are made up of rocks having different seismic velocitiesrseismic waves traveling through then are Often bent too. It is important for the geophysicist and how much bending'takes place. That information portant clues to the velocitrof the rock layers. These clues are extremely useful in decbding what sorts of rocks and what formations Make up the area under investigation. Analysist3f refracted Waveswas the first seismic method employed to explore for-salt domes and the oil which might be associated with them. The technique takes advantage of the fact that the seismic velocity of salt usually is considerably higher than that of thef sedimentary rod( beds surrounding the dome. Stated simply, if the seismic wavetravels from the explosion to the.detector a lot faster than it should if it were traveling through sand or shale, a salt dome and, better still, oil may have been found.

44" Although the word sounds somewhat sirriilar to refraction, reflection refers to a very different wavy quality. If refraction is thought of as bending, reflection is like bouncing. To.understand reflection, it is useful to think of what happens when a ball is thrown againsta wall. A ball thrown directly atwall bounces directly back. If the ball is thrown at an angle, the angle of the bounce is directly related fo the angle at which the ball hits the wall. Seismic reflections behave in a similar fashion. l ite seismic bounce takes place when a wave moving throiigh onkkind of rock hits a layer of another type of rock which has a different seismic velocity. It is the contrast between the Seismic velocitiesin different types of rock that is the "wall." The difference in seism .velocities of two adjacent rock layers, for example a shale and a sandstone, is usually notvery great. So the reflection property of sound waves means both good news and bad news for`the geophysicist. The bad news first. The amount of energy reflected back toward dm the surface is notery great. Typically, less than1 percent of the downgoing energy gets reflected back to the surface at a'single interface becatge so little of the seismic wave is reflected. The geophysicist needs ve.ensitive instruments to reco,rd these weak signals. The good news is thaVhis same lack of reflection means that most

~~48/ONLY ONE SCIENCE it~~

^t . of the energy continues to travelfarther down into the earth for reflec- ion from deeper interfaces. It means that the seismic method can usually he used to map the subsurface through the thousands of feet of rock built up overs_oni of the earth's history all the way to the'base of the earth's crust. This method can also be tire'd to obtain information about the structure of the earth as deep as its core, because the energy that continues downward may also be "refracted" as it travels from one deep layer to another. At deeper levels, however, the data become less precise because of the dissipation of energy.

Noise Seismic waves from an explosion travel in all ditections from their source. The geophysicist wants to detect the refracted and reflected. waves returning to the geophones on the surface. Unfortunately, the source of those waves, the originalxplosion, creates a lot of additional unwanted seismic waves that sometimes overpckwer the much weaker signals corning back from beneath the surface. The geophysicist refers to these unwanted signals as "poise?' The "snow" on a television set is electronic video "noise." This is something like the unwanted sound waves the geophysicist detects. When a channel with a better picture, or signal7 is chosen the noise is reduced. A better antenna, say one on'the roof rather than "rabbit ears" on the set, enhances the ratio of signal-to-poise, and the television picture improves. Fine-tuning might filter out even more of the "snow," improve the signal-to-noise ratio, and result in a better picture. One common way for the geophysicist td improve the signal-to-

, noise ratio is through proper placing of the explosive. Often, the top layer of soil is made up of a variety of materials including rocks and - boulders mixed with sand, decayed organic materials, and other substances. This layer, called-the "weatheredzone," produces copfusing seismic information. By drilling throug4it to the layer below and placing the shot there, the seismic explorer dm minimize the effect of this layer on the signals coming baCk to the'eophones and reduce the Poise.-in effect, etting rid of the "snow" to get a better picture. Many of the signif. improvements in the-accuracy and utility of seismic exploration fo oil and gas ha4been in this area.

Resolution A fifth difficulty for the geophysicist using sound waves to explore the earth's substructure is that the distance from one rock.layer interface to the next is usually quite small, often less than 50 feet. Refleaed from adjacent interfac s, waves traveling at thousands of

~~49 SEISMIC EXPLORATION FogIL AND GAS~~

0t"- rs Eget per secondthousands of miles per houroverlap each otheras they travel back to the geophoneson the surface. The time differences in the arrivals of the waves are very small. Thiscan make it quite difficult to map any single interface. To go back to the example of the television set, reflections from nearby buildingscan sometimes be so strong that the "ghosts" on the tube overlap each other so that the lettering on thescreen cannot be-read or-the picturemay not berecog- nizable at all. In such a case, the signal from the television statidh's transmitter is going not only directly to the antenna but also to nearby buildings where it bounces perhaps to the antennaor perhaps even to another building, all before reaching the set. Because of the differences in distance, a part of the television station's signal takesa little more time to get to the set. The television set has trouble making4ense of such a signal and decodes it as a "ghost" rather thanas the picture originally transmitted. The television viewer usually considers this ghost a minor problem, jiggles the antenna, adjusts.the controls, and decides to "live with it." But geophysicists, charged with the responsibility of making decisions involving hundleds of thousands of dollars, cannot afford to have such an attitude. They need all of the resolution possible Resolution is one of the major problems of seismic research that modern technology has tackled with considerable and continuing success.

Complex Geology In many areas, the rock layers of the earth's substructureare not flat and uniform Forces within the earth have twisted, folded, and faulted layers of rock Waves reflected from these deformedstructures behave in what, from the surface, looks like a nonsensical manner.) Simple plotting of the reflection times from the various rock layers yields a very distorted picture of the complex subsurface- structure. In those areas where the geologic structure is complex, the geophysicist JLeeds a great deal of additional data processing and plotting toget a useful picture. Although this problem has been ch'allenging geophysicists for years and-some significant progres has,been made, it is- getting a great deal of special attention. Some of the most Nepmisin land areas in the United States remaining to be explored for oil andgas are extremely complex.geologically. Each advance in seismic technology that increases the chances of successfully exploring compleR structures underneath the earth's surface enhances the odds of finding crude oil or gas.

50 ONLY ONE SCIENCE HISTORY OF OILN EXPLORATION The Earli.Years In 1845 when Robert Mallet was setting off "artificial earthquakes," oil was already in demand. Petroleum;which literally means iirock-oil,"was being sought and foula. ftor thousands ofye.ars petreleu,krad-been used as medicine,lubricant, and even as a fuel. But the demand flit- this substance had been met by finding surface seeps where the oil either bubbled to the surface in places such as on 4 the Island of Trinidad,or seeped out QE rot* formations where stream beds cut into oil-bearing rock form tions. In the nineteenth century, ' however, whale and fish oils were econting scarce andmore expensive. Improvements in refining me petroleum an attractive substitute for the other fuels. By 18571 e demand for petr_oleunp products, such as kerosene, had grown enough to make the idea of drilling for oil economically attractive. Althougholl had been found as a by.-product -of drilling for water, no one had yet drilled specifically for oil. With the market for oil eanding, however, it was only a matter of time until oil wells wouldake an appearance. The first olltell in the United

The search for oil uas on, and a new industry grewalnwst nightThis photograph was takcti in1865,on the Bentung- hof f farm near -Oil Creek, Pennsylvania, just6years after Drake's successful well was completed ot Titusville Aint`thnPettelleht,11,16{itute

51 SEISMIC EXPLORATION FOR OIL AND GAS ILA Coopers or barrelmakers played an important role in the early days of the development of the oil industry Just 10 years after this photograph togs taken in Omaha, Nebraska areuvd 1890, the United States had produced a billion barrels of oil Amer* Petroleum Institute States was dug in Cuba, NeWYork, in 1857. It was d "dry hole." One year later the Seneca Oil Company of Connecticut began drilling awell on the Hibbard farm near Titusy.ille, Pennsylvania. After drilling 69 feet 6 inches, the company struckpil. (today successful wells are often drilled to depths of 25,000 feet or even more.) The Titusville well was a success. In 1859, its first year ofoperation, it produced 2,000 barrels of crude oil. As the demand for oil greVrthe search for it became more aggressive. Companies that supplied the growing demand reaped for- tunes. The cycle of demand and supply seemed to feed on itself and to grow explosively as new use7for a convenient and cheap source of power were developed. By 1900 the United Stites had produced a total of 1 billion barrels of oil; the second billion was produced by 1909; and the third billion, by 1913. This phenomertal rate of growth was just the beginning. ) Early petroleum production was achieved without the use of geo- On August 20, 1869 aiwin physics. Oil was found by locating wells in places where oil had Drake tompteted the seeped to the surface, by looking for geo,logic structures on the surface first successful oil wellwell near )., Titusville, Pennsylvania Drake, similar to those that had yield it in other locations, or by a number in top hat, is pictured in front of of other means, most of which we not very scientific. Explorers had the well he drilled for the Seneca Oil Company to be willing to gamble, to "wildcat. ' The stakes were high, but so Drake Weil Museum were the rewardi. Many of today's 1gest oil companies trace their

~~53 6EISMIC EXPLORATION FOR OIL AND GAS~~

GO beginnings to this period, but the petroleumindustry would not have survived very long if it had continued todepend on comparatively haphazard approaches to exploration. Fortunately,a large number of people in business were willing to risk their financialresources, and scientists were willing to devote theircareers to developing better ways to find oilIt was natural that investigators tookgeological knowledge and-tried to-m-arry icto lerng=known principlesof-physics to see if there were any.promising means of increasing their abilityto find oil. The problem wasp discovera method for locating likely oil traps without resorting to the impracticalalternative of drilling everywherea kind of'geological blind-man'sbluff. Investigators needed a way to "see" the structure of therock layers beneath the earth's surface. By 1921 two Oklahoma physicistswere ready to try to look beneath the surface with seismicwaves. William Haseman, a physics professor, and John Karcher',one of his former studepts, began attempting to measure seismic velocitiesin rock formations in the area arond,Pronta City, Oklahoma; Theirequipment ancrinethodoloa although crude by more moderQtandards,contained the essential elements used in seismic exPloration today. Dynamiteshots were set off below the surface. The instant ofdetonation, called the "time break," was marked electronically on film running througha photographic film recorder. The seismicwaves were picked up by a microphone detector on the surface, and exact times oftheir travel plotted electrically on the recording film. Calibrationsfor the timing were producedby a tuning fork vibrating at precisely100times per second. Accurate recording of the times it took the seismicwaves to travel from the detonation location, the shotpoint,to the microphone was essenti4 Haseman and Karcher hoped that they wouldbe able to locate types of structures that are often associated with oiltraps. Field tests showed' that their equipmentwas working, their methodology was sound But the real question was, could theirsystem be used to find oil' Despite careful design and theyears they devoted to testing, results were still disappointing. Theproject had to be abandoned Unfortunately, the area their first client hadasked them to explore was not amenable to obtaining good data. Haseman and Karcher'sequip- ment was not.sophisticated enough to make sense'outof the jumbled seismic waves returning to their microphone: Thenature of the geology of the area, not a flaw in theirequipment or methodology, had defeated their attempts. In 1923 another major, unsuccessfulattempt to locate oil-bearing

~~54 pNLy ONE SCIENCE~~

~~61.1 ,~~

structures) was conducted in the United States by a German company, Seismos, Ltd This time the seismic exploration for salt domes focused on the Texas and Louisiana Gulf Coast. The luck of the Seismos crew was as bad as that of Haseman and Karcher.it happens that there are no shallow domes in the part of the Gulf Coast they were searching. There are deeper domes in that area, but the equipment in 1923was not sensitive enough to pick up refractions coming from the greater depths. The following year, however, another Seismoscrew contracted by Gulf Production Company obtained useful refractions from a salt dome in the Texas Gulf Coast regionthe first seismic discovery ofa salt dome that contained oil. ThQ seismic methodwas beginning to prove its potential. In 1926 a seismic crew was hired byGulf Production Company from the Geophysical Research Corporation (GRC),a geophysical contracting company That crew discovered two salt doines within 3 months. Part of the reason for theirsuccess was thatey had placed their detectors farther away from their shotpoint tn had been done k before. They reasoned that by increasing the shooint-to-detector

%ON distance to as much as 6 miles, they would beae to receive refractions from far deeper in the earth and.to collect evidence of deep saleclomeS-. Their results showed that their reasoningwas correct. The equipment was the same, but their new methodology made the leap forward possible. - Another of the GRC's innovations was a new, amplifier that could tune out undesired "ground roll." Ground roll is a seismic wave that starts with the explosion at the shotpoint, travels the surface of the ground directly to the detectors, and drowns out the weaker, but desired, reflections coming back from deep in the earth. Developments such as these enabled GRC to be the only company t during the 1920s to carry out commercial reflection seismographpros- pecting The discoveries they matte for their parentcompany, Amerada Oil, were proving over and over again that the system worked. Because of GRC's outstanding success, Amerada decided to limit theuse of the GRC reflection crews exclusively to Amerada itself and to its subsidiary, Rycade Oil Company. Toward the end of the 1920s GRC employed 70 percent of the world% seismic scientigts, Amerada's decT- 1 sion, therefore, constituted a serious restriction of the development of seismic exploration. Everette De Golyer, president of GRC, disagreed with this monopolistic policy of his parent company In February, 1930 he took an unusual step. De Golyer secretly financed the organization of '".....,

55 SEISMIC EXPLORATION FOR OIL AND GAS Geophysical Service, Incorporated (QS') to do resectionseismograph prospecting. Because the patentson the seismic equipment he needed were held by Amerada, De Golyer's newgroup was forced to develop its own designs and prove them to the oilcompanies. Amerada's successes had made its competitors eager for thenew seismic explora- tibn technology. GSI offered thema way,into the game. Within 1 month GSI had 10-cOritracttiOr as Geopilysical Pksearch Corporation's monopoly was broken. It tookseveraUTears for them to refine the new equipment, but by1938 GSI had 34 cremes workingin North and South America and in the FarEast. GSI was on its way to becoming the largest geophysicalcontracting company in the world. At least 30 new United States seismiccontracting companies were formed during the next 10years or so. Of these, almost all were founded by former GRC and/or GSIemployees. Most of these new companies built instruments similar to thenew design GSI had developed to get around the GRCpatents. However, several of the new contracting companies developed seismicequipment of their own design. One such companywas Petty Geophysical Engineering founded by two brothers, Dabney and ScottPetty. In early 1925 Dabney, a geologist from theTexas Bureau of Economic Geology in Austin, heard about theseismic exploration activities of fhe Seismos crew working ineastern Texas on timberland owned by the Petty family. Dabney decidedthat if he and his brother, Scott, could obtain a seismograph and learnhow to use it, they could go into the consulting business trying to locate promisingsites themselves.

This house is where the Petty brothers developed their prototype at a serstm. detectorIn Scott Petty s too rds twin Seismic Refletnuns), Our rustorrtonths . were tnehardestaraest for we not only had technical problems but financial problems as well Our salaries had ,eased when we quit our tabs Our folks helped out by giving us room and board for free and furnishing us a place to workThe location was ideal fur we found that trolley cars. passing over bumps in the tracks two blocks away, set up helpful vibrations in the earth, which were fairly uniform in nature U StoftPettyarrdUrosvirnrha

~~bl 56 ONLY ONE SCIENCE~~

~~qte ..) /~~

Scutt and Dabney Petty set up the headquarters for some of Dabney asked Scott to read everything he could find about the qfpo- ,ar'y fr,t,,f g ,I- .t,,11, I' Seismos Company, geophysics, and earthquakes. Coiccidentally, Scott st,unients r' this abandoned tarrn house in Brazoria County, had just read about an extremely senstive vacuum tube'"ultramicrom- Texas eter" (a device for measuring very tiny distances) invented by John J. 0 5, qtr. ii,J,IC.,,,,,i,,,1", Dowling of the University of Dublin. Scott ordered one of Dowling's ultramicrometers. Both brothers resigned their jobs and, with $5,000 borrowed from their father, started work at the family home in San Antonio on a very sensitive prototype seismic detector, using the micrometer to measure minute deflections. They first tested the system with the vibrations from trolley cars passing over bumps in the track two blocks away. After some initial field testing and equipment modifications, they leased acreage partly on and partly off a known salt dome. They spent the winter of 1925 experimenting with refractions, reflections, and dynamite shots of all sizes, set off at a variety of distances to obtain .1-

57 'SEISMIC EXPLORATION FOR OIL AND GAS ii variois types of profiles of their dome. They learned how to obtain good salt dome records with as little as 20 pounds of dynamite, rather than the several hundred pounds used on each shot by the Seismos crews, a significant economic advantage. Their first commercial job was to "shoot" a block of family- owned timberland in the southeastern part of Texas. Because of the . results of tests on'their salt dome, they believed they could obtain a reflection directly from the salt by setting a detectorfairly close to the shotpoint. At the Texas site, they obtained evidence of such a reflection. Their excitement at that achievement faded quickly"when they discovered that the evidence waa the result of a loose nut that was vibrating on the detector! Despite that disappointment, they subsequently mapped a dome with a relief (the maximum difference in elevation between the highest and lowest points on the top of the dome) of only 150 feet, a notable achievement. The significance of their success was that, since the dome was about 10,000 feet below the surface of the earth and the relief of the dome was only 150 feet, the, sensitivity of the equipment, especial he recording devices, was fine enough to distinguish between waves arriving only hundredths of a second apart. (Twenty years later that particular dome was drilled and 'prOclueed oil.) For the next 2 years, Dabney and Scott improved their techniques and equipment and had the satisfaction of finding prospects missed by other crews. From 1927 to 1931 the brothers had three refraction crews

~~A tomplete set of refraction eqiiipment that was airlifted to Lenezne1a in 1931 by Petty Geo- physika( Engineering Company~~

~~() "WO nvand 100,~~

~~58 ONLY ONE SCIENCE~~

t65 in Venezuela; in 1930 the brothers sold a complete set of their refraction instruments to Humble Oil and Refining Company. By 1940 Petty had 17 crews in the United States and 7 in foreign countries; their operation was nearly asilarge as GSI's.

The 1940s and 1950s During this period the major oil companies were no longer content to rely completely on the geophysical contractors for crews, equipment, and field expertise. Almost all of,these oil companies formed geophysical departments of their own, staffed them with the best talent they could finci;tand supported active research programs. The oil companies tried to de elop all available geophysical techniques into 0. exploration tools. So e of their effortswere cwite effective. Humble, applying modern technology to scientific principles -discovered more than 100 years before, developed and built the first commerciallyprac- tical gravity meter. The gravity meter measures the average density of subsurface rocks to determine whether older andmore dense rocks have been pushed closer to the surface; it can also determine whether less dense salt has been pushed toward the surface. Using this technique of measuring the earth's gravity at the surface to "see" rock sediments below, that company found 15 prospective sites in Mississippi in 1940. , Many of the geOphysics departments of the major oil companies made attempts to develop practical methods of detecting oil and gas deposits directly by running electric current through rock layers. Today, airborne magnetometers, devices for measuring magnetism in rocks, can map the basement structure of an area from an airplane, sometimes giving clues to the structure of overlying sedimentary rock layers. This technique, based on a per-mile cost, is quite inexpensive and can be used to "screen" large areas for geological structures most likely to be favorable for the accumulation of petroleum. The intimate working knowledge of all aspects of geophysical technology acquired by these and other research groups was of utmost importance in establishing a base for the future of petroleum geology. World War II had both an immediate and a long-term effect on the development of the seismic method. The national effort to develop, new weaponry for use in the land, sea, and air battles of World War II was unprecedented. Geophysical techniques that had been developed to find oil were adapted, refined, and improved to contribute to ways of solving war-related problems. Technology and methodology helped develop better ways of locating enemy submarines (sonar) and artillery

~~r 59 SEISMIC EXPLORATION FOR OIL AND GAS~~

r% DO (echo-ranging). Many of the geophystral laboratories found they had a great deal to offer to the national defense effOrt and devoted at least a portion of their facilities and personnel to the war effort and'to military' projects. Although this deflection of personnel and resources considerably slowed the development of seismic technology until a few yea# after World War II, the technological advances stimulated by war in fields which related only peripherally to seismic exploration were turned to great advantage by geophysicists when the war was over. Wartime advances in the field of electronics led to the postwar development of smaller, lighter, and more sensitive geophones.. The ,------improved geophones could be adapted to new uses. Before the war, for instance, it had been standard practice to use a single geophone for each channel, usually recording four channels of data. A single geophone will pick up and record confusing seismic waves on the surface as well as the unwanted reflectio' ns from depth. New developments made it possible to put multiple geophones togethe on a single trace, or channel, thereby providing a way of cancelling s waves arriving at the different geophones at different times and of reinforcing reflections all arriving at the same time. This technique resulted in seismic recordings with a significantly improved signal-to-noise ratio. Some of the "snow" was gone; the postwar geophysicist was getting a little clearer picture of the earth's substructure. Advances in electronics and in instrumentation stimulated by the war also had an effect on recordings. Before the war, a four-traceA record of seismic waves had been fairly standard. By 1950 recordings using 24 channels were common. More information and more useful seismic clues, therefore, were at the disposal of the oil prospector using the seismic method. In addition to improvements in-the detector atone end of the system and in the recorder at the other, there were advances in the design of the amplifier which connected them. A seismic wave reaching a detector is trans-formed by the detector into a weak electric signal. Before it can be recorded, that signal must be made stronger or amplified, a process not unlike turning up the volume on a radio. In fact, advances in radio and television equipment led to the development of the new generation of amplifiers. Amplifiers were placed on each channel to boost the output of the geophone groups. Each one of those 24 individual trace amplifiers was provided with easily adjustable volume controls and with special high-frequency and low-frequency filters to help cut out various types of unwanted noise. Another "crossover" advance from radio and television technol-

60 ONLY ONE SCIENCE ogy came with the concept of automatic gain control (AGC). The AGC automatically adjusts the volume of a signal to a desired level, even if the strength of that signal varies. Today AGC is best known as a stan- d,ard feature on many home tape recorders; it was also an important 0 advance in the seismic method. In seismic recording AGC is used to adjust the volume on the recorder as the progressively weaker reflecl tions coming from increasingly deeper levels arrive at the detectors on the surface. Such faint echoes, once lost to the geophysicist, are now available for interpretation. Using this new, postwar equiprnent, geophysicists were able.to investigate areas of far more complex geology than had been possible in the past. New areas were opened up for the search for oil. While the new methods did require a lot of slow and tedious work interpreting the growing mound of data recorded from each seismic shot, they were a definite improvement over the prewar methods. By the early 1950s the growing demand for oil had seismic crews searching underwater for oil. A considerable amount of the seismic work was being done in shallow-waterareas such as the swamps and bayous of LouisianarInstead of being put on trucks, the recording equipment was put on boats or barges, or on specially constructed "swamp buggies." The seismic method was proving to be not only curate, but also adaptable to the widened search for oil and gas. Although seismic shooting in open-water areas such as the Gulf o Mexico was underway during this period, offshore exploration was not yet a large percentage of the overall seismic exploration effort. (The first separate cataloging of marine seismic efforts in the Geophysical Actjvity Report of the Society of Exploration Geophysics was not until 19/53.) Strangely enough, because of World War II, more was known about the physical processes involved in underwater explosions than was kno about those that occur in a shothole as part of land exploration. e base of knowledge gaiped from wartime research was to be an impornt factor in the success eiSttle seismic method in marine exploration, and marine explorationw 5Ahe seismic waveof the future. Many of the major oil finds sinc*Se late 1950s have been offshore in open water. Such finds, including the opening o fr1hc vast North Sea fields, were made possible by information gained fro marine seismic exploration. Without those major undersea fields, the world supply of crude oil would have been considerably lower, result- -4 ing in increased economic hardships for all nations depending on imported oil. The description of the state-of-the-art of seismic prospecting in

~~61 SEISMIC EXPLORATION FOR OIL AND GAS~~

63 cthe 1950s may give the impression that the technologyavailable was equal to the exploration task at hand. This might havebeen true had it not been for two' extremely important factors. First, largeareas bf the United States lid the world could not-be exploredwith commercial success because of very,kad noise problems, very complex geology,or very elusive traps. Second, petroldum prospects in the good shooting N areas were becoming increasingly difficult to find. the 1951 seismic-reflection qualitymap of the United States shown below the dimensions of the first problemare quite clear. In the potentially productiveareas of the country, large sections yielded 41.2* either "poor to no-good" reflectionsor only "fair" reflections. - Twenty - five, percent of the area of the United States has littlIsPacpe This map shows the seismic ye- tleLtion quality at the United of producing oil because of its geologic history. Only17 percSti'of the States in 1951 prospective area was judge_ d to give "good" reflections:andonly about Assembly by Paul Lyons 28 percent, "fair."

~~UNITED STATES SEISMIC REFLECTION QUALITY wan' 01 EYPLORATION GEOPHYSICISTS~~

~~11111C3 Or **0 CC,5,,,, 0CS Or Oon ro00 L,,,, (0) bsr mcr,,,, 0, .0.,,,,,04$,. ot~~

~~ONLY ONE SCIENCE~~

a The gliallenge Was plain: first to develop improved techniques that would yield better results in the more than half of the area in the United States where oil might be found but where the complex geologic substructure made seismic exploration impractical; second, to improve techniques for finding petroleum in the good shooting areas where petroleum prospects w.ere becoming more difficult to find. This decline in petroleum prospects can'be seen by looking at the production results of an average seismic crew. In 1934, the golden era of salt dome exploration, a single geophysical crew could expectto find about 14 million 6arrels of oil in a year. But almost 25 years liter, despite all of the significantimprovements in the methodology and technology of seismic exploration for oil, that same crew could expect to find less than 2 million barrels. And they would spend more time, effort, and ,money to find it. While the demand for oil was growing, most of the "easy oil" had already been found. etClearly, there was a great need for a new or an'improved technology that would make it possible to prospect more successfully for oil in thaf large area,of poor reflection quality. Furthermore, data processing techniques were needed which woul0nable the interpreteeto .make use // of the valid reflection informatibn on the seismic recording cot just the smallportion that the geophysicists had time to pick, correct, and plot by crude manual method5

-.. , Modern Developments in It has become common practice to record musicakgroups, rock Seismic Exploration stars, and symphony orchestras alike on multiple sound tracks, with the output of several microphones being recorded separately but synchronously. The advantage of this method is tha after the recording session, the engineer and the producer can adjuste relative volumes, the balance of individual instruments, singers, and orchestra sections on the various tracks. They can experiment with filters on individual sound tracks until they,are satisfied that they have the best possible sound. The same developments that Inaciej5_4eell'ii njque possible for the recording industry also made possible tremendous improvemenfr- in seismic exploration techniques. For geophysicists to be able to handle the problems of noise and resolution more effectively, they had to be able to make multiplg-track field recordings that cou1 l5eplayed qver and over again for analyfkis and could ultiMately belombined in the list possible manner for final

~~4 recording and analysis Magnetic recording on tape had l5ecome a prac- i~~

~~o SEISMIC EXPLORATION FOR OIL AND GAS~~

~~(1 tical reality during World War II. The technology continued to~~

improve in the years following the war,and geophysicists began - experimenting with it. They were looking for a way of recording information from their shooting in the field that would bea significant improvement over the limited methods then In use. One system, de-eloped by Field Research Laboratories of Magnolia Petroleum (now Mobil Oil Company), allowed geophysical crews to record synchronously 13. separate tracks or channels over a wide range of frequencies. By experimenting withmany playbacks of the recording, the geophysicistwas then able to choose the best filter

Y, setting or settings and tune into the precise frequencies that offered the 4 best results in that particular circumstance. The results of seismic explQzation dose with this new generation of the record /playbacksys- tem successfully demonstrated the advantages of the new generation of exploration equipmrt. As a result, the development and the manufacture of equipment for the magnetic recording of seismic-events began to mushroom. Geophysical Service, Incorporated, 4 very large geophysical consulting firm, had, become heavily involved in the design and manufacture of new indumentation during the Kol'ean W ai. That involvement prompted the founding of their subsidiary, Texas Instruments (TI), in 1952. The geophysical company eventually was able to take advantage of TI's pioneering work in semicondUctors by developing new mag- ..rittic recording equipment designed specifically for the seismic

explOrationfor oil. By 1954 TI was ready with a magnetic-disc field- - recording system and an advanced-amplifier. It was the first ina series of new systems and set a new standard for the industry. Other companies rushed into development also. In 1955more than 180 magnetic recording units were either inuse or On order from the 7 manufacturers producing this equipment. These unitswere distributed among 20 percent of all the seismic crews in the world. The magnetic recording revolution was well underway. One very effective use of this new technology was to e from a new method of shooting made possible by this recently developed equipment. It is called the common-depth-point (CIDP) or the commor-reflecting-point (CRP) shooting and processing method. N*, The best way to understand the common-depth-point metd of shooting is to look at an example of it in use. In the figu on Page 65, 24 geophones are spread out in a line 2 miles long on the s4ace of the a tb be. investigated. The diagram shovgl-aypaths of the soundwaves generated by the

64 ONLY ONE SCIENCE al, explosion at shothole "A" (below the weathered layer) as they travel to the first line of reflecting points, representing a rock layer interface, and are reflected to th0 geophones on the surface. The next shot takes place at shothole "B." In preparation for hat shot, the seismic crew will remove the geophone array at the left nd and add another at the right end. They will then explode shothole "B." The results of this shot are illustrated by the second row of reflecting points at the bottom of the diagram. This leapfrog process is repeated P,t,ii..Ica,,7 !this- t,ate..1 21- trak e roll-alcn5prea,i ti ith re- over and over to shothole "L." By the time all 12 shotholes have been '," nh< p,,intt-, a'01,i!ridi- fired, each reflection point will have been covered by 12 diff&ent shots Non The 'at, paths e+ the sowl,i a'ai e ,zenerte,i ht, an ezplo.510h k at .,;10, how are .hr.. r: o

~~Feel 5280 5280~~

~~9ec.,p1",,,nes Shot Holes~~

~~REFLECTING POINTS (Reoresentmg Rock LayerA Interlaces) - B /- ---,r---- \-- D-- E-F - G - H - D JI- - , K - L -~~

(

~~4. 65 SEISMIC EXPLORATION FOR OIL AND GAS~~

K. . 0'1 r's oi 40 from 12 different angles as shown in thefigure below. This type of shooting provides "1,200 percent" or "12-fold" coverage of the subsurface. If shots were spaced twice as far apart, the coverage would be one half, "6-fold," or "600 percent." The amount of coverage required is dependentupon how much the signal-to-noise ratio needs to be increased! And this in turn is dependent on the specific geology of thearea being explored. As mentioned earlier, seismic echoes from each shot travel to the 24 geophone arrays. The output of each array is recorded separately on a 24-trace magnetic recording tape for playback and interpretation later by the geologist. Although the information recorded in CD(' shooting requireka great deal of sophisticated data processing to make

~~.1, .1 .1 - the system work, this method has become one of the most valuable Prad - seismic exploration tools in use today. ' p I f.- z,4 tut;z1- tlq. ;If~~

Feet 5i8: 528,C Geopores Snot HolesX ***

~~Represenhng rock !aye' :ntertaces1~~

~~set~~

~~66 ONLY ONE SCIENCE~~

~~73 T~~

p- The original concept of CDP or CRP shooting, recording, and analysis originated with Harry Mayne of the Petty Geophysical Engineering Company. Mayne conceived the idea when he faced some serious problems on one particular job in 1950. Experience told him 4 that the usual shooting techniques would not give the results needed. The geology of the'area was complex and required a large number of shots. These shots, however, would override the desired seismic reflections with noise. Mayne theorized that the arrangement, which t/ today is called CDP shooting, could provide the information he needed. In 1950, however, there were no practical ways of making a useful recording. Even if he could have made the recordings, the processing of the huge amounts of data generated by the multiple shOts and large numbers of geophone arrays would have been a gigantic task, too / great even to be cottedered using existing technology These problems were enough to make CDP an impossibilityfor a while. Simultane- ously parallel developments in other fields were about to provide answers to the problems of CDP shooting The first such development was in magnetic tape. By 1955 tape technology had advanced to the point where it was practical t9inake field recordings with enough channels to record the output of all the geophone array necessary for the CDP system When Mayne realized that one obstacle had been overcome, he turned his attention to the --4--- problem of processingornassive amounts of data, the remaining stum- bling block to impleritentation of CDP. He made the rounds of all the geophysical instrument manufacturers trying to find a company interested in his problem. Mayne rIallslene way the equipment was developed His description is an insight into the complexities of applying even very promising scientific and technical knowledge in "the f6a1 world I All of the geophysical instrument manufacturers were approached with an outline of the requirements for processing equipment which could perform the necessary correcting, transcribing and summing operations With the exception of one company, the Texas Division of Brush Electronics, no one could understand the need for transcribing the corrected data to another tape and refused to consider making such a machine After considerable negotiation, Brush preNred prelimi- .., nary specifications and submitted a price quotation on a-suitable machine, and Petty entered into a develop- rrient contract in late 1955 Brush was a newcomer in the geophysical instrument business, and not fully I aware of the ltgh standards required by the industry I [Mayne] suspect that their education was a rather painful engineering At1 financial process In any case, after

N 67 SEISMIC EXPLORATION FOR OIL AND GAS . two years elapsed time, considerableengineering trauma, and a price escalation to 21/2 times theoriginal price quotation, they developed quite-a competent machine. Unfortunately, the ge .ysical business had suffereda severe recession in the interim, and Pettyas 4 relatively i small contractor, was dubious aboutthe business wisdom of risking $250,000 (the finalpurchase price) on an .,----Fritried instrument designedto perform a process which had not yet been accepted by theindustry. It was a classic case of the chickenand the egg. We could not demonstrate effectivenessofthe CRP method without the machine, andwe didn't want the machine unless the method provedto be economically viable. All of this against the backdropof a widespread "it will be too expensiveeven if it works" expression of opinion from a large segment of the industry. Consequently, it was withsome relief that we received an inquiry from a major oil company requestingus to relinquish our order for the machineto them. Negotia- tions were completed between the threeparties at the SEG [Society of Exploration Geophysicists]annual meeting in Dallas in 1957, and they assumedour order and subsequently took deliveryon the first machine of its type. It should be gratifyingto all concerned that this particular machine continuedto perform creditably until displaced by digital processingin the mid-sixties. General industry acceptancewas still some time away, and the first clients who supportedthis ugly work were the-,Texas Gulf Producing, Pure Oil and El Paso Natural Gas Companies. Through theinterest of these companies we were able to developa considerable body of experience in a variety ofareas, and confirm our hopes that the method offered significantpotential for data'quality enhancement in most problemareas. t This information was gradually percolatingthrough the industry, and a great deal of interestwas apparent by late 1960 at the SEG annual meeting inpalveston.As evidenced by licensing activity and theproliferation of specialized data processing equipment designedfor the method, interest increased rapidly, from196Q through 1963, and the last major remaining skepticshad finally accepted the method by the end of the 1960's. Mayne goes onto describe the chainreaction set in motion by CDP use. This reactionserves as an example of how a new technology, developed for one specificpurpose, can lead to a.wide variety of .\ changes and improvements affectingan entire field. He writes of this time:

~~Paralleling a part of this period, anothertechnical development was gaining momentum, namelythe digi-~~

~~68 ONLY ONE SCIENCE~~

~~75 it,~~

tal recording and processing revolution: This was also a fortuitous circumstance because the CRP methpd and digital processing were mutually synergistic. The widespread use of the CRP method provided the masses of data required to make digital processing attractive.... One other synergism between apparently unrelated technical developments can be cited between the CRP method and the development of non-dynamite sources, both land and marine. The development of these inexpensive andeffective new sources undoubtedly impacted acceptance of the CRP method by improving its cost-effectiveness. Mayne was referring to the fact that using dynamite as the seismic source for CDP shooting requires a lot of shotholes. Drilling and dynamite already were becoming a very-expensive part of seismic' exploration. Surprisingly CDP, although an expensive process itself, was beginning to reverse this trend and to make possible financial 5,av- ings, at least in some cases. In other cases, the better energy penetration that geophysicists can get ftom exploding dynamite in shotholes . drilled below the weathered zone and the additional information about the thickness and velocity of the weathered layer obtained during the drilling still outweigh the higher cost. Even in the United States where CDP is heavily used, dynamite still remains the most commonly used seismic source on land. However, largely because of the effectiveness of CDP, an earth shaking method developed by Continental Oil Company is running a close second. Called Vibroseis®, the method uses trucks on which large vibrators are mounted; the'vibrators are used as a seismic wave source. At what would be the shothole locations, Vibroseis® creates seismic waves by vibrating the surface of the ground, usually for a few seconds. It then is moved to the next "shotpoint" and theprocess repeated. There is a disadvantage to Yibroseis®. Vibration sources reqtre an even more complicated data processing method to determine the exact travel time of a reflection of the vibration than the method needed to measure the reflection from the very short impulse generated by a dynamite shot. There are circilmstances, however, in which the method's inherent advantages outweigh this consideration. For instance, vibration can be used to search for oil in urban areas such as Los Angeles, where the use of dynternite is prohibited; in addition, the required complex data processing method helps overcome certain types of noise. In addition to Vibroseis®, there are other ingenious nondynamite techniques in use. In one system, a seismic wave is created when a gas , 69 SEISMIC EXPLORATION FOR OIL AND GAS

i II Vtiwseisto a teshmque of t7 Al ,'tiand ,?as tchth is pa/tIcIdarly USeild III areas where the Le.e el,t,tosroes, is not prac- tt,al The vh,,tov-aphs att these .'too qlti,trate this method, Set eta; largt tru,k1, proceed in tandeM along{ a pret,loUsly es- tat,11,hed Path it intervals the ,top and a ,omvuter syrt- b,,,ntzes tits :Ott ewN a' a large steel, ,,t ',ant tiieiueath each t,u, kThe ',,ot :It's the huge elite t t at the ground, and men the t &rates 1,1 urtivn yuirh rr th, other trucks, ..endim< nho,k two the ea,th These t-Mrattont, are de- te, ted ii:f.,phones placed at mter;,als ac,,ss the stir face of the Ltmi The 41b,attous are r,an.pnated to a tomputerrzed dintruck Xed near by/ The V, ks are then lowered to the vound, they ad,ame a short , distance rn rorrnatICr7 and the ..equem e 1% repeated

is exploded inside a container on the surface of theground. fn another system, a heavy weight (approximately 3 tons) is dropped the ground. The purpose of both systems is thesame as that OPthe dynamite explosionsto deliver a strong but sharp impulseto the earth. As with Vibroseis ®, both of these systems offercost advantages in getting the kind of coverage required by CDP shooting. For underwater seismic surveying the variety ofenergy- producing devices used to generate.seismicwaves is even greater. Originally, dynamite was the seisrncw'ave-sourceunderwater, but in the early 1950s it was replaced by nitrocarbonitrate(NCN) because the dynamite explosions were killing toomany fish. However, even if dynamite had not affected aquatic life at all, the growingpopularity of

~~70 ONLY ONE SCIEN '42t~~

71 SEISMIC EXPLORATION FOR OIL AND GAS the underwater CDP method formarine exploration gave thenew nondynamite sources a strong economicadvantage. CDP profiling in water,as on lansd, requires that the energysource be capable of being firedat regular and frequent intervals. Formarine exploration thefiring takes placeat the end of a towline behind the shooting vessel The explosionsmust be uniform and powerful enough to generate seismic pulses that forill travelthrough the water and penetrate through the ocean bottom to Crock layersbeneath. The most widely used of all thenonexplosive sources is the air gun Although air guns had been used originallyin academic studies of the ocean subbotrom and to researchsound transmission through water, geophysicists saw in the airgun an answer to their problem of how to generate seismicwaves under water To create these waves, air.pumps buildup extremely high pressures (as much as 2,000 pounds per square inch)in the submerged gun. The instant this pressurizedair is allowed to rush through the large holes opened in the sides of the airgun, it strikes the surrounding water with enough force to simulatea small underwater explosion. The pulse from this "explosion" spreads ouffrom the source...The de- scending portion hits the bottom andgenerates the seismic wave that will be reflected by underlying rocklayers back to detectors (in this case, the pressure measuring devicesare called "hydrophones") being towed through the water above. One important sidelight to marine searchesfor oil is the difficulty of exploring areas where theocean bottom is composed of a hard,co solidated material, rather than themore usual soft mud. In such a c se, a large portion of the eziergyfrom.the descendingwater pulse is' reflected at the ocean bottom, leavingonly a weak signal to penetate the subsurface At thesame time, the reflected pulse bouncinack up from the hard bottom createshigh-energy noise in the wate Thisin turn creates a very tlifficult signal-to-noiseratio problem whiCh has not yet beep satisfactorily solvedeven by today's highly sophisticated data-processing techniques Certainly,much more attention will be given to this problem as the search for oilin the ocean increases. As mentioned earlier, common-depth-pointshooting and magnetic recording are providinga great deal more data than did the earlier' methods of seismic exploration. From themoment the scientific world became aware of the emergence ofcomputers, geophysicists began thinking up ways to use themto handle seismic data. In 1950 at the Massachusetts Institute ofTechnology, a professor of mathematics, G. P. Wadsworth, anda graduate student, Enders A.

~~72 ONLY ONE SCIENCE~~

~~7 Robinson, began to study ways, of using digital computers to enhance~~

the interpretation of seismographic records The results of their early. work were promising. Within 3 years, with the sponsorship of most of the major oil companies and some of the geophysical contractors, the study expanded. More staff and graduate students were attracted to work on the project In comparison to today's metho ds, the MIT work was primitive, Paper field-recordings had to be hand-digitized for analysis on MIT's "Whirlwind" vacuum-tube computer. The task was not only time-consuming, it was also tedious. . ie. The method of analysis these researchers were using depended heavily on the work of a number of theoretical mathematiciinsAncl., . statisticians. Although that earlier work had provided:a stic5*. , theoretical base for geophysical application, there was Still much left to be done. New theories and techniques applicablepecificallY to seismic waves had to be developed The work went on for 5 years Luckily, as the task continued, computers themselves were being improved. Although they were slow and small in capacity by today's standards, these new machines permitted simpler types of seismic analyses to be performed on a practical basis, replacing some of the more tedious calculations with earlier computer p`rograms There was a fairly gradual transition thrpugh several stages from the use of analog magnetic tape recording and Eninimal processing, to analog recording with digital processing, to digital recording with digital processing By the time the second generation of general-purpose digital computers became available in the mid-1960s, most of the oil companies had switched to digital fieldfrecordings. Some of them started to do data processing themselvese',-both forresearch purposes and to try to improve upon the geophysical contractors' processing results. Core memories of 32,000 wordspr more were available by then, along with multiple times in the microsecond range, so that the processing rapidly became even more sophisticated. Although costs were still high, they were gradually being reduced by the development of new techniques for handling data and by continuing improvements in computer technology About this same time, a promising part of the North Sea was being exploredfor sites likely to produce oil. The processing of data from this area was difficult. The ocean subbottom in this region, with its complex and highly variable geology, stretched all of the state-of-the-art JR digital processing techniques to the limit of their capabilities. The development of velocity analysis methods greatly helped to resolve this impasse. Velocity analysis is the calculation of the speed

~~73 SEISMIC EXPLORATION FOR OIL AND GAS~~

So with which a seismic wave travels througha rock formation. As mentioned earlier, velocity is an extremely importantfactor in seismic exploration not only because itcan be used to determine the depth which the reflections return but also becauseit provides valuable clues to the types of rocks present. In the mid-1930sit became routine to "shoot" wells for velocity. These velocitysurveys were done by lowering a special geophone down the well hole andmeasuring the exact travel times to different depths by shootingat the sface. While this technique gave accurate velocity informationat th well, the data were not necessarily applicable elsewhere. Not until the advent of CDP shooting, digitalprocessing, and the developthent of velocity analysisprograms did it become possible to routinely determine useful velocity informationfrom regular data throughout the entire surveyarea The combination of the seismic method and digital processing made theaccurate surveying of pacts of the bottom of the North Sea possible and led diretftlyto one of the largest oil discoveries of recent years. The effects of thatone discovery on the economies of the participating nationsand on the world supply of oil will ultimately affect almostevery person in the world. In recent years the ability toprocess seismic recordings has continued to improve, and industry relianceon computers has continued to grow Today the United States petroleum industryuses more digital computer power than any other pnvate industry in theworld. Within the petroleum industry seismic dataprocessing is the largest user of that computer capacity. As oil has become increasingly harderto locate, the pressure for techniques that will locate more.subtle configurationsthat may_con tain trapped oil has increased. New instrumentationand new computer programmingve led to the developmtInt of new "high- resolution" techniques. ese methods require even more "number crunching" to extract all pos.le informationfrom seismic field- recordings and to ne theearssubstructure more clearly. The results are impressive. One of these new high-resolutionsystems has demonstrated thateari"picture"a layer as thin as 25 A that is almosta mile below the surface. Such techniquesare expected to prove extremely valuable in finding oil. .A number of prOblems in theuse of the seismic method have been at least partially overcome. Although much researchremains to be done, the seismic system has becomea major tool for the geophysicist* in the,difficult task offinding geological conditionscapable of entrap- ping oil and gas.

~~74 ONLY ONE SCIENCE~~

~~.. 81 I~~

OTHER USES OF THE SEISMIC METHOD While its role in the search for oil and gas is probably the most widely publicized, the seismic system has made significant contributions to other areas as well. In the mining industry, the primary traditional applications of the seismic method have been to map the depth of rock layers down to bedrock or to explore for ancient channel deposits that may contain .,...... 0 uranium. More recently, high-resolution techniques for seismic exploration have been modified to map coal seams. The spread of the seismic method into the mining field has been somewhat limited by a lack of appreciation for seismic techniques on the part of mining geologists. But acceptance of the seismic method is growing as it continues to prove itself as a cost:effective means of collecting valuable geological information. The increasing value of minerals has made the seismic method (a very expensive way to locate minerals) more economically feasible. The seismic method is also being used to study potentially important geothermal power sites. The structural information collected about geothermal power sources deep inside the earth will greatly facilitate the development of these energy resources. Some of the refinements in the technology used in the search for oil have also been adapted to study earthquakes, just of some principles of studying earthquakes have h &lped geophysicists locate hydrocarbons. The scientists_who_study earthquakes_and_their behavior, causes, and even their prediction are working with techniques that are familiar to the petroleumbgeophysicist. There are also similarities in the equipment that is used. The instruments rely on the same scientific principles; the major difference is in the source of the "bang." In seismic exploration the vibrations are set off when and where the investigator wishes. With earthquakes, movements of layers of the earth itself generate the vibrations. The earliest known "seismograph" was invented in China about 2,000 years ago. Choko, the inventor, designed a sculpture; it was a-circle of dragon heads, each with a copper ball in its mouth, surrounded by a circle of fro-gs. When the earth shook, the copper ball fell from a dragon's mouth into the waiting mouth of a frog ( page 76). As quaint as this arrangement may seem, the Chinese reported that this device detected an earthquake about 250 miles away. Scientists who stlidy

~~75 SEISMIC EXPLORATION FOR OIL AND GAS .- . 0) This mOdel ot Choko's graph was made In I950 by. re terrin,z to an old Chinese draw mg _ Nah.,"4415t-orce-tfus,um in Tokyo, japan~~

. earthquakes today are interested in knoWing not only thatan earth- ;5- quake-took"rilace but also how strong it.was, how'long it lasted,the degree of horizontal and vertical movements of the earth's crust;where in the earth's crust it was generated, the kinds of motions that took place, and when and Avhere thenext one mighWcur. The answers to such questions are helping to reduce loss of life and damagetoprqp- erty from one of the most deslructivefe...1.6rcei innature. Americans tend to think olearthqUa,s as a problem of some very "sinrcific parts of the country, especially of California.They forget,that Cities such as Boston, St. Louil, andCharleston have also been shaken A by severe tremors. Because ea?thquakescan occur in.a number of locations, studies using thekrinciples of the seismic method have been useful in choosing locations for structures for whichthe preventicin of , damage fromearthcinakes is particularly criticalAuchstructures include nuclear plants, dams, public beuildings, skyscrapers,and bridges.. '* Vibration detection devices similar in principleto those usedin oil ,exploration have been modified to record how different building es. materials arid various arch ect ral designs reactto simulated earth- e quakes. Scientists are using the information from thesetests to engineer.structure,s that will withstand the tremors of quakeSf- guarding the lives of,people who live'anc work in thosesections of the

76 ONLY ON,E SCIENCE V . a 3 0 Scientists are n'ot the only people fascinatedby.theTesults of the work in the Piedmont region. The sedimentarystrata under the overthrust rocks turn outtaisa bevery much like rock layers that often contain petroleum of one form or another. In addition;the dense crystalline rock thrust over them makesan almost ideal seal for oil and as which might be stored in porous sedimentary` rockbelow. In fact, as a result of COCORP, several oil companies havesent seismic crews to the Piedmont area in the hopes of finding oiland gas.Phlyexploration wells can show whether there is commercial oilor natural gas available within those very deep sedimentarystrata of the Piedmont. The findings of COCORP have suggested that portions ofthe United States previously considered barrenmay contain badly needed oil and gas. Whether searching for oil at depths of25,000 feet or looking for clues to the geologic history of the earth at depths ofmany miles, the scientific principles involvedare similar. On a COCORP expedition there may be university professors and experienced oilexplorers working side by side, each able to makean important contribution of kniNedge and experience towardthe solution of the problems at hand Theurriversity community and the oil-industryhave leai-ne-d that the principles of seismic exploration providea versatile and accurate tool for seeing beneath the surface of the earth. Theseismic method owes much to basic science, theroots of what penetrate the history of science from the development of thesilicon chip back to Newton, Galileo, and Da Vinci. It is today successfullyand increasingly repaying that debt

~~4.4~~

~~78 ONLY ONE SCIENCE~~

Sz5 sc, Scientists are n'ot the only people fascinatedby.theTesults of the work in the Piedmont region. The sedimentarystrata under the overthrust rocks turn outtaisa bevery much like rock layers that often contain petroleum of one form or another. In addition;the dense crystalline rock thrust over them makesan almost ideal seal for oil and as which might be stored in porous sedimentary` rockbelow. In fact, as a result of COCORP, several oil companies havesent seismic crews to the Piedmont area in the hopes of finding oiland gas.Phlyexploration wells can show whether there is commercial oilor natural gas available within those very deep sedimentarystrata of the Piedmont. The findings of COCORP have suggested that portions ofthe United States previously considered barrenmay contain badly needed oil and gas. Whether searching for oil at depths of25,000 feet or looking for clues to the geologic history of the earth at depths ofmany miles, the scientific principles involvedare similar. On a COCORP expedition there may be university professors and experienced oilexplorers working side by side, each able to makean important contribution of kniNedge and experience towardthe solution of the problems at hand Theurriversity community and the oil-industryhave leai-ne-d that the principles of seismic exploration providea versatile and accurate tool for seeing beneath the surface of the earth. Theseismic method owes much to basic science, theroots of what penetrate the history of science from the development of thesilicon chip back to Newton, Galileo, and Da Vinci. It is today successfullyand increasingly repaying that debt

~~4.4~~

~~78 ONLY ONE SCIENCE~~

~~Sz5 sc, (~~

~~its~~

~~%,, \, N . -~~

~~Ftr I;'z~~

~~t"?. %Ars;x1.\\A~~

. . efhni 101 %14NeIt d brat tards with hole. p14,1c vt them tIt tidal:fed to the tabu la thm ote,,Stt, Ft (araytve,e lewd tiugpac him' he ',Yet et- .opel t,l, ,dtttq the United State. (13,7%14, t# 144Q0 I

~~80 ONLY ONE SCIENCE~~

~~1 Survey Research,aripl Opinioh Polls~~

In 1888 James Bryce, a British lawyer and Member of Parliament, R noted a "new force" in the world, "conspicuous only since governments began to be popular." That rorce was public opinion. Lord Bryce believed that whether a country is despotically goyerned or free, "both are generally ruled by opinion." But in free countries people* their supremacy, treating their rulers as theitagents, while irrdespo\tic countries the habit of submission prevails. Bryce felt that a stage in the evolution of opinion, frOm unconscious and passive to cor*ioutind active, could be reached "if the will of the majority of the citilTifs) were to become ascertainable at all times" without having to pass through a body of representatives or even through a cumbersome voting machinery. But, of course, the "meckaniCal difficulties...of working such a method ofvemment are obvious." Bryce.yvrote: - How is to will of the majority to be ascertained except by counting votes? How, without the greatest incon- venience, can votes be frequently taken on all the chief questions that arise? No country has yet surmounted 'these inconveniences today, some 90years since Bryce wrote The American_Commonwealih, not only can public opinion be frequerttly ascertained but the thnol- ogy also exists to determine it instantaneously. Whether or notis A latter capability has strengthened democratic systems is a matter of

controversy. "N Survey research is a science that provides a variety of tools for dis- covering, manipulating, and organizing data in many fields, both academic and commercial. Most often, the general public comes into contact with just one aspect of survey researchthe public opinion poll. 1 Public opinion polling4s not an easy term to define, although the t ...--

~~81 SUK4Y, RESEARCH AND OPINION POLLS~~

~~a *.4~~

method is used in all social sciences today and inmost related profes sions Some define and use the term narrowly to referto a method of data collection, while others think of opinion pollingas an analytical tool It has been used'for evaluation of the past and for prediction of the future. It can provide accurate information about thepresent; however, when it deals with the past, the data are sifted through and censored by human memory which often issiot'tioo reliable. is It has been estimated that $4 billion are spent eachyear inithe United States on all kinds of survey research,an industry that employs millions of people directly or indirectly and has hadan impact on every Amencan The capi4 investment for survey research in staff, training; equipment, supervision, and interviewing is large Anarmy of trained interviewersis constantly on call to undertake projects The public which reads newspapers and magazines, watches television, or listens to the radio is very familiar with the political opinion poll Somewhat less frequently, the public alsosees or hears new analyses of various social, political, or econom issues thatare based on "climate of opinion" surveys. It would be misleading, however, to be1' ve thaf survey research is used primarily tci.provide the press wit sessment of the standing of political figures With the electorate, o Wrmine how the publicf'eels ul the issues of the day. S search is also an important too lof the United States Census Bureau, whichuses it to collect information on population characteristic's, employmentstatis- qcs;cornmercial indices, and other social indicators. Market and advertising research would be impossible without sample surveys. Radio and television iatings, sociological and anthropological studies that, among other things, influence the distribution of welfare funds, educational progress reports; economic analyses, the climate for mill-. a tary recruiting efforts, and public health informationall use the data gathered in survey research. In short, whenever it isnecessary to gather data about people in the aggregate and their relationshipsto one group or another, survey research is one of the major scientific tools that makes this possible. Sur\;7ey research has many arrtecedents, Asa means of gathering data'about the demographic.characteristics of people ina socialsetting (that is, their sex, education, income, orage distributions), surveys go back through history. Siirveys were done in the past because society heeded the facts to take appropria4'e actions, and theyare done today for the same reason. Careful descriptive surveys were conducted in the late eighteenth

~~82 ONLY ONE SCIENCE~~

4-4- / 'and early nineteenth centuries in Europe and especially in England. John Howard gathered a systematic and objective body of facts in the 1770s that led to prison reform in England..Henry Lytton Bulwer and Frederic Le Play described French social and economic conditions in the early nineteenth century, and Henry Mayhew gave detailed descrip- tions of various segments of London society in the mid-1880s. The studies of these men were not based on scientific samplings nor were they a complete census, but they_did point to the value of social surveys, even when the samples on which they were basedwere not truly representative of the population.

re . The application of representative sampling to social surveys followed the industrial revolution and urbanization. With larger numbers of people living in one place, it was particularly impoetant, when laws and regulations were proposed or made, to know as rapidly as` possible who would be affected and how. Before 1880, when the population of the United States was less than 40 million, census enumeration of the entire population took 10 to 20 months. After 1880, with a population of over 50 million, more

...... _census-takers wereOded, and the count was completed in 1 month: Still, the analysise the census of 1880 took 7 years. Herman Hollerith, an employee of the Census Bureau, helped to halve this time by sug- gesting that a card with holes punched in it that had been uspd 'for some time by weavers to produce patterns in cloth be adapted to the tabulation process. (Hollerith's work is also mentioned in the chapter, "Computers and Semiconductors.") But even the use of the card was too slow for many needs. Sam- pling of representative parts of the whole was adopted by several individuals and governments as a compromise between accuracy and speed or, alternatively, economy. Sampling made it possible to get information fastel- and cheaper than through a census of the entire Herman Hdtterith population. In England and France, for example, sampling of sorts was (1860-1929) ckL, IBM Archives used al the mid-eighteenth centuzy even before regular censuses were introclucecMy the end of the nineteenth century, the pressures for fast. and cheap population data had reached the point where sampling for demographic information had become .a necessity.

Opinion sampling was a different matter altogether. Only a ._ p. long-shot gambler would try to predict the outcoMe of an election on the basis of a straw or "trial" vote, bufthiS is just what the editors of the Harrisburg Pennsylvaniandpd the Raleigh Star did. They askcd their readers to indicate how they would vote in the elections of 1824 and based their predictions on this unscientific approach. Their chances of

~~83 , SURVEY RESEARCH AND OPINION POLLS.~~

~~8:)~~

~~_ ' -7'4/ 'tr'~~

~~P.41*.CULTY.A.737410,1411Pfl...41A1,,VII;;;A-C-4 4111ElialOuntierrADMVORIflloVIEVSMCIWIECILOIC&ARMstRy, AND IANUALTUREEL~~

~~Her/mall 11,4!, rah kMat;Imes, used intheteoo were eat irrt d th+Tont er en ti#I 4 Ilirli all 4 10430 )800 xrrninnto iaciawai.- Sc. sokI~~

~~85 SURVEY RESEARCH AND OPINION POLLS~~

u a correct prediction were not verygood.!though the Raleigh Star correctly anticipated that Andrew Jacks 6nwas in the lead, he received .# nowhere near the 80 percent vote he received.in thqstraw poll. Nevertheless, y the end of the nineteenthcentury, such straw polls had become crimon in both.the dailynewspapers and magazines. . With a very few exceptions, scientific opinion pollingdid not start until the 1930s when it was introduced in boththe marketing and the broadcasting industty. Samptinghas its basis in themathematical theory of probability. Once sampling's scientificvalidity was understood, its social, political, and commercial valuebecame obvious. Prom then on, social scientists were encouragedto work on improving the sample reliability and to expand itsarea of usefulness by developing or adapting statistical tools'that permittedmore sophisticated-analysis. World War II gave another strong impetusto the development of survey research The Surveys Division of the Office of War Informa- tion conducted surveys ofoverseas civilian morale, while the Research Branch of the Division of Information and Educationin the War Department helped to train, indoctrinate;And evaluatethe military forces Since thenthe development and availability of highspeed computers to collate the data, more sophisticated mathematical techniques to analyze what the data mean, and vastlymore advanced collection techniques have allowed the science of sample-surveysto Opetate with a much firmer scientific base and with inLYeasing reliability. The applications of survey researcti have been varied and broad.Most people mee with it more frequently in connection with public pinionsur- veyBut the same principles of sampling applyto the many other uses of sur. ey researchwhether-it is in the collectionof data for econom- Ics, legislation, education, anthropology,or any other field that involves examining the distribution of characteristicsin an aggregate.

MEASUREMENT Asking questions is fundamental to socialsurveys. Statisticians of the early nineteenth century used questionsas an instrument with which to measure social phenomena. But they alsofound that; despite t. carefll definition and classification of theevent (or 'iiil?jgct) to be mea- sureorsubsequent observations of theevent (or object) were not neces: sarily an exact replica but only dreasonable facsimile.The prototype of this facsimile, called the "average man,"was described in 1848 by Adolphe Quetelet, a Belgian mathematician,astronomer, and social

~~ONLY ONE SCIENCE . 4~~

~~fj"~~

0 scientist. He wrote that individual observations will show deviations from the average, but the greater the deviation the less frequently such observations will occur, distributing themselves around the average in the form of a "normallbell-shaped) distribution. Qu_eteler believed that the concept of normal distribtition could be applied to all aspects of the study of society. Mathematics, he decided, was the only way to turn the study of society into a science, because mathematics can focus on what is common to objects. If one could generalize about a branch of knowledge to the point where one could classify things under common headings, it would be possible to reach an advanced level of that science. Adolphe Quetelet But to be able to classify things under common headings, uniform (1796 -1874) Lttlani at Coniress methods of data collection, tabulation, and present of findings are needed. He applied this precept in this measure.- -nt of the p cal characteristics of people (for example, theationships between the heights and weights), and he insisted thone could do the same for moral and intellectual characteristics. 40. The English scientist Francis Galton, best known for his pioneer- 1/4 _ ing work on heredity, carried Quetelet's theories one step further. He was concerned about the way people were often classified as either . dttbeautiful, honest or dishonest. In his study of eugenics, he fotind that Quetelet was right; whatever a person's characteristics weretheSc were normally distributed around a mean of the charac, teristics..1n other words, people had a "central tendency" of beautyor honesty. The further one deviated from that tendency, the fewer people theretoere who cdtild be said to be that extreme in the characteristic. To discover this, Galion had to devise fairly unambiguous measuring instruments or questionnaires and a rating scale. The first mailing of a'survey quegionnaire Was initiated in the 1780s by a weelt4 Scottish landoiwner, writer, and parliamentarian by the name of John Sinclair. He sent out 881 questfonnaires to the Scot- tish clergy. The questionnaire was divided into parts dealing with geography andnaturaistoi-yof the parish; age, sex, and'occupations of thepulation; and farming and minerals.,He also introduced *anothsurvey research techciciue,the.follbw,-up 'Otter, still used with mail questionnaires. It has been found thatiollow -up letters ar crucial to improving the representativgness of a mailsample. . One other branch in the lineage of survey research is.the concept of attitude. In 1848 Quetelet tried to relate various popillatiOn characteristics, such as age, to the probability that an individual might,,have criminal tendencies, He called it a criminal penchant or propen0,ty. He

~~87 SURVEY RESEARCH AND OPINION POLLS~~

~~9kJ .."-L.. .2~~

~~...-~~

. . 2 .._.,1 0 said there was a distinction betweenan "apparent propensity".and a The former is what appeared on the surface; it could v 2- be observegl and calculated. The "realpropensity" is the underlying reason for the "apparent propensity"; it'cannot be observed,but only inferred from a person's behavior. Realpropensities were what were tater referred to as attitudes. -. One of the first attempts to measure attit'udeSthrough a large- scale survey Was by Adolf Levenstein,a self-educated German laborer. Between 1907 and 1911, hesent out 8,000 questionnaires to miners, steel Workers, and textile workers, askingthem about their opinions on issdes suchas their political and economic hopes and aspirations, their religfods views, satisfaction with theirwork, their drinking habits, and their cultural pursuits. Germansociologis.t Max Weber prevailedupon Levenstein to publish his findings andsuggested ways for him to tabu- late and analyze them. Weberwas stimulated to think about the application of survey research to the building of theoryin the social sciences.

The Discipline of the Much survey 'research is done to determine theextent of a partic- , Scientific Method ular type of behavior (such as.readinga particular magazine), attitude (such as favoring a bill to permit srnoking ofmarijuana), or characteristic (such as country of origin4.fora liven pOpulation. Most newspaper -.. polls and pOlitical opinionsurveys are of.these types. The findings lead nowhere beyond providingan answer to a question. If the question is irrelevant, the answer is irrelevant. Kurthermore, ifthe question is a fixed alternative type of question', offering the respondenta choire-of only a few possible answers, the-resuffsO alhe survey are limited by the imagination and ability of the survey designer and such otherthings as the nature of the problem Asan example, imnine that researchers were living in a culture in which it is Aieved that illnesIs caused bya person's beinepossessed -by devils. If theywere-to design a study to determine the cause of a particular type of ill.ne'sq they wouldtry to analyze the kinds of devils thatwee in pee* wlithwith different types of illness Because they already "kne;v" theigeiceralcause-of illness, the researchers would have no need to examineany other possible causes; people do not ask questions that do notoccur to them. . , 0, The concept of the controlled study is basic toexperin}entation- i (A controlled study in the social sciences isan e5

Structured Versus Interviewing for surveys in the nineteenth century was largely Unstructured unstructured. Questions could boexplained by interviewers in their Interviewing own words and could be expanded. Questions were not always phrased the same way in face-to-face interviews. To collect demographic data, highly structured questionnaires were not crucial. But in searching for something that is not physically observable, such as a person's attitudes, the only thing that could be held constant was the question. If the questions were not the same, the answers were not really measurable. Unstructured questioning has it uses, however. Unstructured surveys are exploratory. An example: the United States exhibit in the 1967 World's Fair in Montreal had not been a suc-

~~..,..------ft1/4......~~

~~89 SURVEY RESEARCH AND OPINION POLLS~~

. . 93 I ; cess. So for the fair in Osaka, Japan, a means to m ke th exhibit more appealing had to be found. It was decided toappro the problem through an unstructured survey. A set cif questionswas prepared, and trained psychologists were put to work interviewing Japanesewho hail been in the United States for no longer than 6 months. Theywere asked what were the most interesting things they had experienced when they first arrived in the United States, abort what thingsthy had written home, and what specifics they wished friendsin Japan could see. The unstructured interviews produceda set of ideasothat * were:then worked into a structured (standardized) quAiprinaireto be , used in a survey in Japan. -The structuredsurvey established the degrte of interest in particular types of exhibits, while the unstructuredstir- vey prOvided the more creative ideas for incl sion in the structured survey All forms of unstructuredinterviewing,however,*require later desitandardization by the rfsearcher for thepurposes of measurement ', i4,111ere.searcher mustState theideas thatare common to.the vaii- respo ;: although each respondera.may have used different Sys of stabg there. _ k ,.. , . y 4 '2. S cture., interviews permit measurement "ofreactions tpan unvaryin tion. If an average response tole given questions Iscon- ceived, then it should be possible to measute.theamount that:a given populakipn varied from thavverateresponse. The stirkulus is the 9uestion; the response is the answer to the qtestion. -.. ,.-- .. . . - Plumbing the Human One problem with survey research is.that it mup,t deal with"self- Mind report." Survey researchers 41ave long beenaware of the fact that human memory is imperfect and that people eitherconsciously Or unconsciously tend to distort-theirresponses. Even factual questions, such as, "Do -you read comic strip's regularly,sometimes, or not at all?" 1 oi, "How often do yoU read 'editorials inyour newspaper?' are not _ answered accurately, sometimes quite unconsciously:Thetendency on the part of most people is to provide the socially desirableanswer. Some of the social desirability is eliminated in mailor self- administered surveys ..One tInique that has been developed to counteract the tendency is totroduce the question with: "Most' s. people, as you know, occasionally takethings that donot belong to- them ..." or "Some people believe. .while others believe...." Respon- dents can thus align themselves withone group ortflother ar,t:ot feel like misfits. Another problem of-Self-report is knownas response set. Some people have a consistent tendency to agree witheverything. Others

~~90 ONLY-bNE SCIENCE~~

k .tend to be very negativeand disagree with everything or to respond to every question on a scale by picking the extreme choices. A survey research technique that has beep developed to reduce the effect of agreement or disagreeinent is to wrife the questions that Night pose problems in,a positive vein for half of the sample and negatively for the other half. Thus, half the sample is asked: "Would you say you agree or disagree with the statement: 'Most people are dishonest?"' The other half of the sample would be asked the same question, but "honest" would be substituted for "dishonest." For the most Pars, survey research depends on verbal responses. Exceptions are such involuntary measures as galvanic skict responses on which the use of lie detectors is based or measuring thelcontractions and dilations of the pupil of the eye, a technique yThich is used in some market studies. Fortunately, people usually tend to tell ,the truth- because of moral persuasion, because they find it easier than telling a lie, or because they are not involved enough to want to lie. It is when people are motivated to disguise the truth that the survey researcher must be most careful. No satisfactory method has yet been devised for assessing the probability that a respondent is distorting answers, although there are some techniques to reduce nonresponse or distorted responses in the case of sensitive or difficult to answer questions.

Projective Techniques . Most people are reluctant to answer sensitive questions about and Role Playing themselves To overcome this reluctance, survey researchers use projective techniques which ask respondents how they think ".others" or "most people" would answer a given question. By projecting the respondents' own behavior to "others," they avoid embarrassment but reveal their own attikities and behavio'r. During the first decade of the twentieth century, the Swiss' psychiatrist Carl Jung began developing oneforn of projective testing word associatioestudies. Another major contributor to projective tests was the Swiss psychiatrist, Hermann Rorschach. His tests are "often referred to as the "ink-blot" test because subjects are asked to react to'a spot on a piece of paper which resembles an itnic blot. The Rorschach test was the basis of yet another form of projective testing occasionally used in survey research. This other kind of tek the Thematic Apperception Test, was first developed by Haryard psychol- 9gisi Henry A. Murray in 1935. It involves showing respondentsa picture and asking them fo use their imagination and, through free associ---_-_ ation, lo answer questions about it. As an example, in a*village in South-east Asia an interviewer shows respoindents a picture of a fatnily

~~91 -SURVEY RESEARCH AND OPINION POLLS ,~~

4* sitting around a radio set. "To what program are they listening?" the interviewer asks The respondents know that most.people listen to for- bidden Communist broadcasts, but the villagers would be foolish to admit to a stranger that they do so. Yet this is amairt-believe picture. There is no problem So a respondent says the fam is listening to a Communist broadcast. Unfortunately, most respondents approach a survey interi,c.iew-as though it were an intelligence test. Philip E. Converse,a survey

research specialist at the Survey Research Center of the University of, Michigan, found that the average respondent, whenf aced withan opinion question dealing with an unfamiliar topic4s reluctant tosay, "I have no opinionI have never thought about those things before you came and never will again after you leave " Instead, an opinion will be ventured. In a 1956 study Converse tried to relieve the respondents who might not have an opinion of any compulsion to. answer by intro, ducing the question with. "Of course, different things are important to' different people, so we don't expect everyone to have an opinion about .all of these If you don't have an opinion, just tell me that "in spite of this, follow-up studies and analysis showed that only an additional 27 percent gave no opinion upon this.special inVitation. It was assumed, based on previous experience, that about 9 percent would have opted far a "don't know" response anyway ThAt left 64 percent who chose to offer an opinion Of these, 45 percent could be shown by comparing and analyzing their responses in follow-kip interviews to have no opinion The 1956 study by Converse and other research have shown that only around one in five members of the public has an informed opinion on matters dealing with public affairs. Unless an opinion is grounded in strongly based attitudes, it is not a reliable guide to future bahavior.

Question Wording The problem of how to word a questio'n to obtain the desired information is a difficult one, and-there is no fool-proof Pule to guide r- the survey researcher As Humpty Dumpty put it in Through the Looking Glass and What Alice Found There "When I use a word...it means just what `% choose it to mean, neither more nor less." This, of course, isan exag- geration because culturally and socially shared meanings of words exist Bui the fact that one.can maneuver people into responding positively to a most serious question affecting everyone in the population was illustrated by American sckial.Psychologist Hadley Cantril. Using the responses to ..surveys conducted by the American Institute of Public

~~92 ONLY ONE SCIENCE~~

~~1 100%~~

SO FAR AS YOU, PERSONALLY, ARE ABOUT /Vas' CONCERNED, 00 YOU TmE USANDNoTFAF HAS CONE TOO FAR INHELPI.NG SOME PEOPLE SAY THAT IFTHE -US BRITAIN , OR NOT CAR ENOUGH' GOES ON HELPING ENGLAND, YES 6- 2 4 - 41 4-Cw8ICO -76% GERMANY MAY START A WAR AGAINST OUR COUNTRY DO YOU THIVK WE SHOULD °CONTINUE TO IF IT APPEARS CERTAIN TO YOU HEL-P ENGLAND, EVENIF wE RUN THAT BRITAIN WILL BE OEFEATED 73% THIS P,SA 5 - 6 - 41 UNLESS WE USE PART OF OUR NAVY TO PROTECT SHIPS GOING TO BRITAIN WOULD YOU FAVOR OR OPPOSE, SUCH CONVOYS' 5 - 6 - 1 US co , INTOWARWOULD YOU PREFER TO HAVE WHICH or THESE TWO THINGS DO THE US GO INTO THE WAR, YOU THINK 15 MORE IMPORTANT HELP RATHER THAN SEEBRITAIN SUR- FOR THE U5TO TRY TO DO - BRITAIN RENDER .TO GERMANY'5- 6-41 TO KEEP OUT Or WAR OURSELVES, OftTO HELPBRITAIN, EVEN AT 58 z THE RISK Or GETTING INTO WAR' DO YOU THINK THE US NAVY 5-29- 41 SHOULD BE USED TO CONVOY es' -`SHIPS CARRYING WAR MATERIALS TO paITAINt 6 - 24- 41 ,ROOSEVELTAND OUR LEADING MILITARY EXPERTS SAY THAT 59% BRITAIN WILL BE DEFEATED UNLESS WE GO INITO THE WAR IN THE NEAR PAVOR. 4"" FUTURE, WOULD YOU FAVOR, OR OPPOSE, GOING INTO THE WAR WITHIN A rEW DAYS'5 - 20- 49

~~41.~~

~~IF YOU wERE ASKEDTO VOTE TO- DAY ON THE OuE5TiON Or THE U3 ENTERING THE WAR NOW AGAINST .0 GERMANY AND ITALY, HOW WOULD YOU 'VOTE 70 GO INTO THE WAR r a 22X- NOW, OR TO STAY OUT Or THE WAR' 6- 24-41~~

~~SHOULD YES US ENTER THE WAR NOW, 18% 9 - 17 - 4t~~

~~77IVC~~

~~7.'17'77 °mum SHOULD THE U S GO INTO THE WAR -JJ r NOW ANp SENDAN ARMY TO I1 ro I '7. '7 .1 EUROPE? 9- 17 -41. ," " ,,,,7,..~~

Now ()pillion Varies With Contowne los involved During May through September, 1941, ,1111,e ,bowed that between 8 percent and 78 percent of Americans favored involvement nt World War 11 What made the difference 14,41,S whether they were asked bluntly,Shall we send an army to Europe'" or an indirect question su.h as,(fie you think the U 5 has gone toth_far m help- Butrun or not far 'enough ''' .

~~93' SURVEY RESEARCH AND OPINION POLLS~~

9;) pinion and the Office of Public Opinion-Research, just before the nited States entered World War IL Cantril showed that from 8per cent to 78 percent of Americans favored involvement in the war in Europe. /

Note in the figure on page 93 that differences in responses hinge o whether the question is asked directly, by implication-ror by a combination of direct and implied statements. , Over the years, certain rules for question construction have evolved. One rule that has been recognized for a long time is.that for most purposes questions should be asked in the most simple and economic form possible. The words used should Se understood by the person with the least education in the sample. Sometimes the question is a concealed two-part questiop, as in the following qu?ry froma Congressman:

Health care costs hav increased substantially in recent years, to the extent that adequate health care has become difficult for many people. Should the Federal government consider a National Health Insurance program to cover all Americans?

The respondent who is for a national health insuranceprogram, but not for one that covers all Americans, is ina quandary. The ques- - tion should have been simplified by being split into two 'separateques- tions. Split question techniques have the disadvantage, however, of lengthening the questionnaire.

~~Leading questions are another-favorite of those whoare more interested in the public relatiOns effect of theirsurvey than in the climate of opinion. This example is from another questionnaire:~~

~~Are you in favor of a tax cut, even though it might 1 eii to a greater Federal deficit?~~

~~. . Some questions are simply ambiguous. They mean different things to different people, and the survey researcher carrn of tell whichwas meant, as in the following question:~~

~~Do you believe t. ai gun control can solve the crime problerfi?~~

94. ONLY ONE SCIENCE OP The meaning of "gun controE may range for some people from simple registration to coViscation of all firearms; for others "solve" may mean either reduce or eradicate. Adding all of the responses to this question is like adding-apples and oranges. The split-ballot method was developed in 1940 by the American Institute of Public Opinion to test the effect of wording differences.! The method, which is often used by market researchers, as well, involves using two versions of a question with different halves of the sample of respondents., The following is an example that shows the emotional contekt of the italicized words: f A Would you favoradding a law to*the Constitution to prevent any President of the U S from serving a third term

~~B Would you favor changing uthe Cbnstitution totprevent any President of the U S from serving a third term?~~

~~.1.0 Percent Percent . .1 Response forResponse for Percent .. Responses Adding' 'Changing Difference~~

~~Yes 36 26 10 No 50 65 15 No Opinion 14 9 5 ( *Not emphasized in original survey -,~~

~~The order in which questions are.asked also can have* an important effect on the response, as in the following:~~

~~A. Do you believe that anybody shoube allowed to speak on any subject she or he wis to, or do you think there are times when free spe@ch, should be prohibited?~~

~~B. Do you believe in free speech to the extent of allowing Communists and extreme rightists to bold meet- ings and to express their views in the community?~~

95 , SURVEY RESEARCH AND OPINION POLLS tlecauseIle researcher has little controlowl' the order in which questions are read in the case of self-adnlinis ered questionnaires (as in a mail survey), questions that mey be affe by their sequence are difficult to ask, and must be masked or camflaged. In the modern questionnaire there are essentially two kinds of questions: open-ended and closed-ended. In open-ended questions, respondents fill in their responses (or an interviewer dues it for them) in their ownwoids. This-is useful when the researcher hasno idea what the anslers might-be. It is used extensively in a pilot or preliini- nary study when information is being collected about scope or issues. Open-ended questions.have their drawbacks, hoWever. If the interview is being conducted by an interviewer, it takes a fast writeror a tape recorder to take down everything the respondent says. If the interviewer summarizes, thg danger of injectingpersonal interpreta- %ay tions arises. If the questionnairels-self-administered, some respon- 'dents may tend to skip open-ended questions because they are too muchbotRer. If everything is taken down verbatim, there is also the problem ofsoding ater. Someonemust decide how to classify the answers under a 1ited number of categories, otherwise answersca .not be reduced to numbers for purposes of comparison. Closed-ended questions, also called fixed-alternative, multiple- choice, or cafeteria questions, precode the responses. This means that the researcher lists the choices for the fespOndent to considerthe queition: For example: Compared with 5 years ago, would you say your atti-* tude toward your locfl hospital has Become betterdas remained about the same, or has become worse? .,

Closed-ended questions have the advantage of fOcusingall respondents' attention on the same alternatives. Theresponses can thus be more easily tabulated. On the.other haft{, the alternatives Presented force the respondent into ansutificiil mold. The true opinions of the population maylie elsewhiere. George H. Gallup, in 1947, develatid a procedure that he called the quintamensional design to overcomesome of the disadvantages of open-ended and closed-ended questions..This method begins Withan open 'ended general knowledge question to determine whether respondents are aware of the question. Then it is followed by another open elided cwstion in which respondents are asked for an opinion George Gallup -about the issues in their own wordgf Next comesa question with alter- (1901- ) ' natives ons e specific part of the issue, and respondents choose ' Tht Gallup 5tgantzattorri among them. is is followed by a qpestion asking "Why?" to permit

~~96 ONLY ON&SCIENCE~~

1 U respondents to elaborate on their choice. finally, there is a cloSed.- ended intensity question to determine how strongly the'-views are held. Because of the possibility'tha't trends in opinion reay be studied,, at I some future time, questions are often worded in eicactly the same way as they were in some past survey. Any changes in worf ing would make it difficult to determine whether it was the Chan e in wording or a change in attitudes that influenced the results of a surrey. When responses to questions are tabulated, it should be remembered that they are respeses to words that have only approximately the same meaning for the various responctnts This is one of the reasons why sample surveys do not necessarily reflect the "true Qpinions" of the population. ... Another reason is thatsurveys reflect a situation at a given time; . yet they are often projected into the future. When asked how they will vote tomorrow, people May oblige aid tell the interviewer how they think they will vote tomorrow. But Often implicit in their response is_ the !Naming, "I hope you wonThold me to it." Survey research may be an important guide to the future, but that does not necessarily make it ,..a go& guide Yet it is a principle, bothin everyday life and in scientific endeavors, that the best evidence available shOuld be the gtirde.\Survey research ha&yro;;ided increasingly better evidence on which to base decisions on certain topics.

SAMPLING Its Origins In 1912,4!group of city officials with a problemk Reading, Eng- land approached Arthur Bowley, a part-time lecturer in economics and statistics at the local University Ex4ension College. They were con-

cerned about the condition of the working class in the city. But Read- . Mg could not raise the money for a socialsurvey of the magnitude of earlier, more sweeping studies. Bowley was just thopersonwackle this prlDblem. Although his education had been in mathematics at Cambridge, his interests were in economics and social issues. He had heard athe great debates at the Internation-ar Statistical Institute con- ventions at Bern in 1895 and Budapest in INV1 on the "representative x. method." He had also heard that Anders N. Kaier, atief of the Norwegia Bureau of Statistics, had used "representative sampling" in gathering abOr statistics.in 1891. So he undertook to use the limited

~~'funds available to prepare a report on "H'elihood and Poverty" in . Reading/Itwas a landmark study for sampling techniques.~~

~~97 SURVEY RESEARCH AND QPINION POLLS r'~~

4 IhReading, it was.economy that forced theuse of sampling. In 1923 in Japan it was the need for speed that precipated itsuse. An earthquake in Tokyo destroyed part of the tabulation of tlfeJapanese census. To get, the results out by the following year-7by,age,sex, size of households, and other demographic characteristicsJapaneseoffi- cials took a sample of 11,000 by picking every thousandthhousehold of the 11 million in the census. When the tabulationsof the eritire , census were later-completed and compared with the sample results, it was found that the sample had provided very accurate data. . ) George H. Gallup; who started a conimercial polling organization in the early 1930s, showed thatas long as a sample was representative, it did not have to be very large to reflect consistent resultThus a survey of public opinion on prohibition provided the followingresults:

~~Percent Percent Percent, Fav4ing Opposing with No Prohibition Prohibition Opinion~~

~~First.sample of442 respondents 31 62 - 7~~

~~Adding second sample of442,thus ) 584 1.makirtg respondents 29 63 8 Third sample of443,making1,327 respondents 30 63 7 Fourth sample of1,238making'2,585 respondent? 31 61~~

~~Fifth sample of2,670,making5;255 respondents 33 59 Sixth sample of2,098,making8,253 respondents 450 Seventh sample of 4441, making 12,494 I respondents 32 61 7~~

True, it is not known what the distribution of opinin regarding prohibition was in the, Population as a whole. But (hereis no reason to believe that further sampling--to the, point where the total population is'approachedwould have caused percentagesto fluctuate any more. It took a long tine for the laws of probability and sampling, which had been usedin connection with natural and physicalobservations, to be applied to social phenomena. Jakob Bernoulli,a seventeenth- century Swiss mathematician, physicist, and theologian, demonstiated in 1713 thiat the colors of-.-fample of marbles drawn froman urn would

~~98 .4 ONLY 9NE SCIENCE~~

I U be approximately proportionate to the colors of all the marbles in the urn if the marbles in the urn.Vvev randomly distributed. This illustrated empirically, or through experiment, What Dutch mathematici;n Christian Huygens anti French mathematician's Pierre de`Fermat and Blaise Pascal hackfipred out theoretically 50 years earlier. (Huygens' contribution to seismic explaration is mentioned in the chapter on that subject.) Bernoulli stressed that a substantial number of observations would have to be made for the sample to be accurate.- Bernoulli suggested,ticat sampling might be applied to economics. About 100 years later, in the early 18.00s, a French mathematician and astronomer, the Marquiscle Laplace, tried to figure.out the probability that jurors might err in arriving at a verdict. His compatriot mathematician, Simeon Denis Poisson, worked on the same problem a-rid, in 41837, developdd the bases for the'application of statistics to the social, sciences. But, it was not until the end of the nineteenth century that' British scientist Francis Galton began to apply stat. tics\to biological problems. * it had been the common wisdom that "like b gets like." Galton set abdu4 proving this axiom empirically and statistically. To .do this, he measured and. weighed fathers and sons. He found a high correlation

between the heights and weights of fathers and sons, biit he also . discovefed that fathers who were very much taller than the average tended to have shorter sons, while shorter than average fathers .tended to have taller sons. He called this "regression" to the mean (or average). Bowley, the man from Reading, pointed out in 1912 that what could be done with biological pfoblems could also be done with-socio- rbgical ones. He stressed that, while the number of variablesencoun- tered whendealing with human beings may be unlimited, what is important is the variability of single attributes of these human beings, such as how theiopinions vary on.a given topic or the extent to which. their opinions differ from the average. Actually it was this measure- ment4nd observation of variance that first led to the study of probability in physical observations. Around the middle of the seventeenth century,, a Frenchman by the name of Chevalier de Mere called on Blaise Pascal to help him improve the odds on his gambling. Fascinated with the ch4llenge, Pas- cal worked on the problem and checkecN out with his friend, Pierre de ThFermat. One of the first things Pascal did was to toss coins. He then- worked out how many chances he had of getting a given number of heads and tails in any number of trials t.

~~SURVEY RESEARCH AND OPINION POLLS~~

~~u i. . . Being.a mathematician, Fermat immediatelysawtht the nymber of timers each chance is likely tooccur could be figured out mathemati- cally, thus: f~~

~~Tossing a single coin one can geta Head or a Tail 3~~

~~With two coins, one can get~~

~~With three coins, onecan get~~

~~100 ONL'JNE SCIENCE~~

~~196 4~~

Pascal prepared A triangle of chances based on up to 10 coins that looped Something like this: I '''..-tirnberof Coins Sum of Chances 1 . , 11 2' 2' 1 2 1 4 3 1 3 ,3 1 8 4 7 1 4 6 4 1 = 16 5 1 5 16 10 5 1 32 6 1 6 15 20 15.6 1 64 7 1 7 21 35 35 21 7 1 ' 128 8 18 28 4,6 70 56 2,0 8 1 . 256 9 1 936 84 126 126 fl.4 .ip 9 1 512 10 110 45120 210 252 210 120 4510 1 1,024

~~1098 7 6 4 32 10'~~

He read the last fine of his triangle as follows: If 10 coins are tossed once or 1 coin 10 times, there is 1 chance ins1,024 of getting alLheads or tails, 210 chances in 1,024 of getting either 6 heads and 4 tails, or 4 heads and 6 tails; etc. Graphing this, he found the ollowing kind of figure that show the decreasing probability of getting mostly heads or mostly tails:

~~Number of Chances in 1,024 N.. 0~~

~~' 252~~

~~210~~

~~120 45 10~~

~~1 1 2 3 '4 5 6 7 8 9 10 Number of Heads (or Tails) 'Of~~

~~io1 SURVEY RESEARCH ANDOPINIOAPOLLS~~

~~1 C-~~

'" Abraham DeMoivre, a French mathematician refugee in England, in the early 1700s smoothed the resultingcurve by increasing thp number of cases (or coins) to a very large number, and in the latterpart

of tilt century the French astronomer-mathematician Marquis de , Lap lke introduced further improvements. In the early nineteenthcentury, the German mathematician and astronomer, Karl Friedrich Gauss, used the Laplacian curve todevelop the conepttof standard deviation and of standard error from the mean. To provideany d,ta, however, that are susceptible tojnathematical manipulation,researchers have to agree on just what it is theyare measuring. Observation of nature had shown thatno two things, atleast among biological and social phenomena, were exactly alike. In other words, generalizations here are theoretically impossible. But if minor differences could be ignored, it was argued, categoriesor classes of 'things and ideas that are approximately alike could be lumped- together. Since, however,. nothing is an exact replica of anythingelse, if something is named, sa-y,-"leaf" and thought ofas a ideal, every other leaf after that deviates from the ideal. Statisticians referto the deviation as "error." The second leaf is not identical, butit is still close enough to be called a leaf. What Gauss didwas to defineand calculate the average error that is likely to be friundin anythingin nature. called the averageerror a standard deviation, whichis a form of a age deviation from the mean (or average) of all observedcases. An example of how mathematicians calculate thi'; varianceor deviation from the ideal mi0it help to clarify the theorybehind it. If the term "weekly income".i6 used at the example,a sample of this income might be the weekly paycheckover a period of 5 weeks, thus:

~~First Week $120.00, Second Week 80.00 Third Week 70.00 Fourth W,eek 100.00. Fifth Week 80.00~~

The total income for the 5 weeks is $450, and theaverage, or mean, income is $450 divided by 5, or $90. Note that the Mean isan ideal or "e representative weekly income. None of the weeks in this particular \ 4

~~if 102 ONLY ONE SCIENCE J~~

~~1 i.- a j~~

sample shciwg that amountyEach week is in- "error" to some extent- that is, if $90 were theideal representative of the term "weekly --\ / income," Gauss figured out how to define how much each,was in error. He also noted that theean, by definition, is a midpoint. If the deviations from the mean, the pluses and minuses, were to be added, they would total zero. To eliminate.the zero, he squared the deviations, since this would not affect the relative distance of the deviations'but would take care of.alLthe minus signs, thus:

~~Squared DeviatiOn - Deviation Observation Mean from Mean from Mean i' . . . t t First Week 120 Q0 +30 +900 Second W'ee'k 80 90 ' -10 +100~~

~~, . _, Third ,"\,..ek 70 90 -20 +400 Fourth Week 100 90 *I 10 +100 ,Fifth Week 80 90' '-10 +100~~

~~Total 450 450 0 1,600 ,~~

The average squared deviation from the mean is 1,600 divided by 5, or 320, and this Gauss called a variance. If the variance were "unsquaree or the square root taken, the result would be about 18. He called the number calculated in this way the standard deviation. When Gauss examined the Laplacian curve (which has since bell rename,d the Gaussian curve), he deduced its mathematical form and noted that the curve had the shape of a bell. He figured out that about 68 percent of the area under that curve wars within one standard deviation from the mean, 95 percent of the area was within two standard deviations; and better than 99 percent of the area was within three standard deviations from the mean. To illustrate his discovery, the measurem6nt of the results of intelligence quotient (IQ) rests can be used.-For this example, average IQ can be given a rating of 100. Obvi- ously, Pot everyone has an average IC):. Some are above average-and soe are below. How much peopleyQ deviates from the average can be ca uted by subtracting 100 from each person's IQ rating and, /workin out the standard deviation. If it is discovered that the standard deviation is 15, according to Gauss' approach about 7 out of 10 people would have IQs that measure no less than 85 ano4 no more than 115that is, one standard deviation-(15) on each side of the mean (100). 4f

~~103 SURVEY RESEARCH AND OPINION POLLS a. 4~~

~~'t I 'et~~

Furthermore, 95 percent of people would beno more than two standarddeviations (15 multipliedtby2,-ch 30)removed from the mean, putting them between 70 and 130 in IQ, and 99 percent of people./ would be no more than three standard deviations from themean '(45), piecing them between 55 and 145on the-IQ scale. A great deal of work has been done to determine the optiMal number of responses needed to,assure accurateresponses to `survey questions. It has been found that even if a survey is repeated a number ,- , Of times Using different samples of the Same size, theproportions for a particular question will varyfrom sampleto samplebut not by much. These changes irpresponse rates from sampleto sample are known as the sampling error. To halve theerror, a sample four times that size is needed. Thi,s,is wherea compromise must be made between accuracy and ecohomyt Two questions must be asked. The first is, 'how much error can be tolerated? Ina survey dealing with attitudes toward a polaical,candidate, a fairly-largeamount of sampling error may be acceptable On the other han&if a survey were made to detei- mine whether a drug will have harmful' side effects,a much smaller error would be tolerable. The second question is, how importantis itto be certain that this sampling error will not be ex-eeededOne standarddeviation covers close to 70 percent of the curve. Actually,some 32 percent of the sample would lie outside this area. Thus, if the samplingerror tolerated is 3 percentage points, and a survey showed 60 percent favoringa candidate, the true population percentage might be anywhere between57 anct763 percent. It wouldbe about 68 percent certain that thiserror of 3 percentage points is not exceeded. Two standard deviations would provide over 95 percent confidenceor only a 5 percent chance of the error being greater,-and tliree times the standard deviation yields better ,than 99 percent assurance.

Populations and Imagine 3-carload of grain. To determine low much impurity is in Sampling Frames the grain, a random handful could be taken to count the number of foreign particle's in.that sample, assuming that the entire sample is well mixed. If the handful contained 1,000 individual seeds and 10 of the particles counted were foreigoparticles, that wouldrepresent a ratio of 10 to 1,000, or 1 percent. There is no need to examine the entire carload. 'This one handfulwould give a fairly good estimate of the incidence of impurities. X 1. The real world, however, is not that homogeneous; populations are not uniform. In the 1870s the German statistician Ald economist

~~104 ONLY*ONE SCIENCE,~~

~~110 - 0~~

Wilhelm Lexis developed a theory for cluster sampling. That method involved drawing samples from population clusters which tended to have more homogeneity internally, Or within a cluster, than there was between the clusters. He devised methods for drawing subsaMples from each or the clusters so that the final sample was representative of the entire population. A major problem with sampling is defining a population so that- there is no doubt whether a given unit or element is or is not a nvinber of thabpopulation. Unless the population can be defined precisely, samples may not be repiesentativ9. Having defined a populationfor example, all citizens who are registers 'to vote and are not traveling abroad or in jail or in a hospitalit is necessary to find a sampling frame from which4o draw 4 the sample. This sampling frame, a list or map of all the units in the population, must be comprehensive and up-to-date. A telephone directory, often used as a sampling frame, illustrates many of the problems of defining the population. If the population is defined as "telephone owners," adirectory does not include unlisted numbers, nor does it include all names of individuals who share a single telephone.' Moreover, it includes many names of people who have.moved or whose numbers have been changed as well as business telephone numbers that are not identifiable with a particular person or persons.. If theylare so identifiable, that person often is listed again at his or her resiAnce giving the person more than one chance of being'selected In, short, defining the population unambiguously''and finding a sampling frame to fit the definitioh are difficult and important challenges to sampling and survey research. Failure to meet these challenges has cau4ed some massive errors in the past. For 20 years, the Literary Digest had been conducting presidential election surveys b'ased on mailing-Its it had started collecting in 1895. In, 1936, 10 million ballots based on Ml" ephone and car registration listings were mailed out and 2,367,523 straw votes were returned. These telephone subscribers and automobile owners gave 54 percent of their vote to Alf Landon and 41 percent of their vote to-Franklin D. Roosevelt. The electorate, on the other hand, gave Roosevelt over 62 percent of their votes. Obviously, the Literary Digest voters, in spite of their numbers, were not representative of the electorate, The magazine's straw votes were discontinued that yearas was`the Literary Digest itself Two approaches to achieving representativeness have bey attempted. quota sampling and probability sampling. Quota sampling

105 SURVEY RESEARCH AND OPINION POLLS involves dividing the population into relevantgroups based on factors such as age, educational level,sex, geographiCal location, and nee. Quotas are set far each of these groups inproportion to their preva- lence in the population. Put simply, if20 percent of the population has a college education and 40 percent a high school education, while anOther 40 percent has less thana high school education, then those are . the proportions by 'Which the sample is, drawn. Quota sampling has number of shortcomings.Accurate and up-to-date figures on population demographimareessential to setting the correct quotas. But,more importantly, therelevknt characteristics for a particular survey must be knownsomething'hat is normally better known after the survey than before F ore, if the interviewers are given free rein as to whomey include in the sample to make up the qtlota, there is always the chanceof picking people who are most like the interviewersthat is, more congenial. Quotasam- pling, therefore, is beins usedless and lessin serious surveys, especially in view of the fact that theerror cannot be estimated when it is used The probability that characteristicswill fall within a calculated margin of error is based on the assumption ofrandomness in the choice of subjects, andpure randomnesscannot be assumed in quota sampling. There artwo approaches to probability sampling.The purest and best is based,on a listing of-the entire population,from which a random sample is drAn k usinga table of computer-generated random Numbers which indicate the elementor persort to bpicked. Numbered listings of theopulation foruse as a sampling frame are hard to obtain or do not exis. So a second method, based-on the geographicallocation of theelemeng of thppoptilation, is usedmost of the time Area sampling,as it is called, 'mightuse a map over which a grid is placed, marking out cells thatare picked randomly. Within each cell or cluster thus-chosen, everyonecan be interviewed, or the clusters can be enlargedansfsdbclusters can be picked forinclusion in the sample. The problems of selecting respondents; of dealingwith refusals, "not-at-homes'," wrong addresses, andwrong identifications; and of making substitutions plague survey researchersto the present day. One basic'rule of sampling is that fora sample to be representative of the population from which it is drawn, itmust be truly random, which means that every element in the population must havean equal or a known chance of -,eNing selected, andevery combination of elements must have an equal br a known chance of being selectedanalways difficiaEgndsr sometimes impossible conditionto ensure.

~~106 ONLY ONE SCIENCE~~

~~112 - r~~

While sampling theory has changed little intfepast 30 or 40 years, the methods of securing and analyzing the data and of reaching the rctpondent have.ehanged significantly. Even more significant is the difseovery of Ohat will and what will not work. In'the 1920s and early-1930s, mail survey's were popular. But by the mid-11930s, commercial and academic pollsters had decided that the personal interview was superior, although far more expensive The Voice ofAmerica has used radidannouncements where access to the pdpulation is difficult, if not impossible, especially in sparsely populated Afridan countries ilhete callon listeners to write in, giving some information about themselves, in exchange, for which they will participate in a drawing fora prize. One obvious advantage of mail surveys is relatively low cost. Another is that the interviewer does not affect the responses,by influenckg the answers. The digadvanta:ges are numerous. the respondents arebself-selected, that is, the respondents decide whether they will or will not return the questionnaire. The bias is toward people in the upper educational and socioeconomic levels, who are more likely that/ those in the lower levels to answer their mail. The sequence in which the questions are answered cannot be controlled,andat may influence the responses. There is no way of telling wh ther the ques- 4 tionnaire was filled out by the addressee or by some ofr person. And, most important, it is not possible to'tell hoW the vi of those who didinot reply differ from those who did. For many years it was the consensus that the only truly scientific jurveys were those'conducted through personal interviews. Their advantages remain numerous. They are still the best way to exercise full control over the sample, although reaching the selected individuals may be difficult at times and maytake a number of callbacks. They are certainly the best method for long interviews. Because more than 94 percent of households in the United States have telephone service, telephone surveys have gained in acceptance since the 1930s and 19405. Today, it is possible to reach almost all segments of the population and to reach them quickly and relatively cheaply. In most cases, the int'iews must be kept shortperhapg no longer than 20 minutes it reckt studies suggest that once the 4 respondent is on the tele hone, tl4Tength of the interview ceases to be a major problem.-the biest problem is with complex and long questions, which are hard t minister over the telephone. Questions that involve the ranking of alternatives, or even selection among a large number of chokes, simply cannot be administered.

~~107 SURVEY RESEARCH AND OPINION POLLS~~

~~A . 0~~

11t) Panel studies involverecurrent interviews with thesame sample of individuals. Theyare frequently used for consumer studies and for broadcast ratings. But they havealso been used tQ,study.how the electorate makes up its mind,as in the seminal Er,ie County, Ohio, surveys by Paul F Lazarsfeld, Bernard . Berelson, and Hazel Gaudet in1940 Panels are most useful instudies of trends and causes of change,but they suffer from a high drop-outrate and from the dangers of contamination. Contamination, which isextremely difficult to determine, occurs when responses to later interviewsare influenced by condition- ing caused by earlier question;those who are exposed to thequestions are more likely than others to notice relevantitems in the mass media or to discuss thetopic with friends. Broadcast ratings are often done throughpanels. The A C Niel-, sen Company has used a meter that is attachedto television or radio sets and checked periodically Nielsenand the American Research Bureau also use a diary method, lighichinvolves leaving diaries witha panel of subjects who fill themout as they listen to a radioor watch television The problem with metersis that they merely record whether the set ois on and to what channel it is tuned,not who, if any'- one, is listening or viewing And it is impossibleto ensure that diaries are kept when memories Ore still fresh Coincidental telephone interviewsare done by C. E. Hooper, Ync., Trendex, The American ResearchBureau, and others. They involve calling at the time the program is being broadcast to determine how- many people are listening and whd these petkpleare. One problem is that the person most interested in theprogram is least likely to answer theshone. "Day recall" telephoneinterviews conducted by the Sindlinger Company, among others,overcome this problem. People are asked which programs they heardthe previous day. The weakness is the shortness of many ofpple'smemory. Personal interviews covering the previous 24 hours, conducted byPulse, Inc., have the same shortcomings In recent years, electronic devices havebeen introduced both to assist in the interviewingprocess and to obtain.instantaneous feedback from viewers of a televisionprogram. One such polling device received wide publicity in July of 1979 whenit was tested for instant feedback from 6,000 to 8,000 viewers ofa presidential speech to the nation. The polling system works in conjunctionwith the Warner Communica- tions Company's QUBE cable televisionservice in Columbus, Ohio. (QUBE is alsadiscussed in thechapter, "Synthetic Fibers."),Its30,000 subscribers have control boxeson their television sets with numbered

108 ONLY ONE SCIENCE Viewers of QUBE cable television in Columbus, Ohio are able to Lummunuate their re- buttons which register at the studio. Percentages reflecting'subscriber sponses dirpetly to the studio responses are calculated instantaneouslythus making it possible for thereby pro:Wing instant feed - buck to questions asked by the the audience to provide and to receive immediate feedback. Because station subscribers are representative of nobody but themselves and tend QUBE toward the upper socioeconomic levels and because the respondents who choose to participate in any particular survey are-not even representative of the subscribers, results of such polls are no more than interesting at this point... Ultimately, the objective of a sample survey is to make meaning- ful statements about the population from which the sample is drawn. Theoretically, one can say that proportions found in a random sample will be the same as those to be found in the population as a whole, within calculable margins of error. S.

~~109 SURVEY RESEARCH AND OPINION POLLS v~~

goliticqlPollini Irk 1432 George Gallup conducteda survey for his mother-in-law, who was-running for Secretaryof State in Iowa. Shewon the election and th s became the firstwomn Secretary of State in Iowa. The first paid - o ey commissioned by a politician was undertakenin 1946, when Jaco K. Javits had theRoper organization undertakea survey of the 21st Congressional District ofNew York.-But, by the mid-1970s, three out of four congressionalcandidates were using professional pollsters, at least to write theirquestionnaires or to design theirsam- ples, if not to conduct the entiresurvey.

The niforrnation gatheredmost frequently in the first poll ina political survey deals with issues importantto the electorate. Atleast as many questions are asked to determinethe relativetstanding of candidates Somewhat fewersurveys-examine the candidate's image and the strengths and weaknesses ofthe opponent, and a relatively smaller number of surveys also checkon the effectiveness of the variousjnedia with the electorate. Thecosts have already run into the hundreds of thousands of dollars and continueto rise Since 1960 when pollster LouisHarrieconducted the first private poll for a presidential candidate(John F. Kennedy), no candidate for national office has run withoutthe help of surveys. One estimate indicated that by the mid-1960s,85 percent of the winning senatorialcan- didates and more than half ofthe elected members of Congress had used polls , The accuracy of reputable pollshas acreased tremendously since the fiasco of 1948, when almost allpollsters predicted that Thomas E. Dewey would defeat Harry" S. Trumanin the presidentiaLelection. The average combined error of final congressional andpresidential election polls c4 the Gallup organizationfrom 1950 througH\1978 has been about one-half of 1 percent. But not all pollsters have provedto be as accurate as the Gallup ipolls, and some are UnsOupulousin their use of survey research. Charles W. Roll, Jr. an Albert H.Cantril wrote in Polls: Their Use artd Misuse that when the New' York DailyNews showed Charles Goodellrun- t( ning a poor third in the Senate'raCe,his stiff was contacted bya polling Xrm which asked if Goodellwould be interested in purchasinga poll that showed himto be ahead: Another firm, bidding fora senatorial candidate's business, offered to providetwo surveys: one for fund-- raising and publicitypurposes and another to show the true standing of the candidate v Dan Nimmo reports th /T candidates , themselves often misuse stir- vey findings. In Novemberof1967 Lyndon 8.Johnson's staff

110 ONLY ONE SCIENCE I lated eprivate poll showing the President leading all Contenders In a "bellwether" area. Archibald M. Crossley, who had conducted the survey, revealed that his findings had been distortedthat the survey had, in fact, been done iiia traditionally Democratic county in New Hampshire. )

Government Surveys , Governments haye always had a need for information about the people they governed, and, in one way or another, they have always 1 managed to collect this information. The concentration of large numbers of peoplein urban areas during the industrial revolution increased this need and called for more detail. Rising imrpigralion led to the collection of city-by-city dafa on country of originVith the census of 1850; beginning in 1870, educational reforms required detailed dalta on occupation and education The need for accurate population,data kept increasing, and more and more studies were done. But these were generally hasty and poorly-conceived reportskhat ignored the availability of sampling fechniques being developed atI ihe

44.... time. . i . It was not until the census in 1940 thatthehe United States Census .. Bureau added a 5 percent sample survey of hodseholds to include questions on employment and income. Since then, the Census Bureau e." has been adding topics until today it has quarterly rotating panels to estimate population characteristics such-as social and occupational mobility, the labor force, income levels, ownership of major appli- k antes, housing plans, and consumer buying habits. During World War II, the Research Branch, Information and Edu- cation Division of the U.S. Army, did about 3studies oncolie one - half million soldiers. These studies helped inolicy decisions affecting troop morale, training, racial desegregation, and demobilization. They also provided a generation of studtmts and scholars with a gold mine vf ditaand.methodological models for studies in social and communication research. After World War II, a number of other goyernment agencies added survey operations, including the Bureau of Labor Statistics of the Department of Labor, the Congressional Research Service of the Library of Congress, and the International Communication Agency, which conducts opinion and media surveys ih many,foreign countries. These surveys are used in foreign policy decision-making in evaluating the impact of its own media and cultural activities. Government -4-onsored and government-conducted surveys provide d ta' for many commercial end social uses. The surveys are essen- MI. \ 111 SURVEY RESEARCH AND OPINION POLLS

~~'WO~~

~~,s. r~~

)

. tial to the proper functioning of a modern society. But the datathey generate must be accurate a goal which can be achieved only if they area basedon scientific survey practice. Ina4curate data that masquerade as accurate and scientificare not only misleading and useless but are also potentially dangerous.

Marketing and Media Among the early names associated with business and marketing Surveys research was that of Archibald M. Crossley. Startingas research, manager of J. H. Cross advertising agency in Philadelphia after gradu- ating from Princeton, he later became assistant director of research for the ill-fatedLiterary Digestpolls. In1926 he set up his own business as' a market researcher, using surveys asa business aid for newspapers, advertising agencies, and corporations. Hewas one of the first to sample public reactions to-brand names, and when radio broaciciSling became widespread, he pioneered insurveys of radio advertising impact and listening habits. One of the most widespreaduses of survey research today is for rating radio and television programs. Com- mercial ratings of network radio programswere first undertaken by the C. E fHooper Company in 1935. Then, the A. C. Nielsen Company, founded in the early1940s, invented its mechanical recording device, the audimeter, thatwas attached to radio receivers to monitor the station and theamount of time a set was tuned in to it. Nielsen bought the Hooper Companyin 1950. By the mid.7,1970s Nielsen was offeringan "instantaneous Niel- sen" service by connecting some 1,200 setsto a central computer for overnight tabulationofaudiences.. Pulse, Inc., entered the broadcqt rating businessin the mid-1940s using the personal interview method. In 1952, the American I Research Bureau added the diary method (and latertelepherraoincidentals and recall surveys). Today the American Research Bureauis called Arbi- tron Ot.ber companies, such as Trendex and Sindlinger andat least 50 more, have been doing both national and local rating surveys since the e"/ 1950s TheAccuracy and the methods of the broadcast ratingsgenerate a great deal of controversy. In 1961 they were the subject of hearings by Representative Oren Hartis' subcommitteeof the House Interstate and Foreign Commerce Committee, which reported mat;short comings some of which have now been corrected.' Today, much of the estimated $4 billion that isspent annually on survey research goes into consumer and media audience resdarch. Cor- porate market'surveys help stirnAte what types of productsconsu- mers are likely to buy. Time ses analyses in estigate past buying u.

~~N 112 ONLY ONE SCIENCE~~

~~1 1j I~~

behavior to try to predict future patterns of buying. Market testing-is done with the help ofsample surveys to determine whether a new product has a chance'of capturing a large enough segment of the marRet to make marketing it worthwhile. Market share analyses help a manufacturer determine how successfulproduct is in competition with similar products of other manufactuipes. Packaging regliirch uses -t surveys to study the most appealing ways to package a product.

ANALYSIS Probably the most progress in survey research has been in the sophistication of the analysis of data. Computers provided a major impetus to this development. The earliest survey researchers were satisfied with mere frequency distribution, percentages, and a simple

breakdown of the data by demographic characteristics. In short,they reported who gave what answers, what proportion gave each answer, and what the proportions were in each town, age group, sex,Incqine level, and so forth. While the analysis of the data is far more sophisti-, cated than it wag 30 years ago, validation of the findings still needs improvement. Not much attention was paid to the validity of the find- 'ings until the 1930s, when American psychologist Samuel A. Stouffer began to ask whether a relationship existed between what was measured and the intention of the investigator. The validation of survey data is difficult. The data are valid only if they measure specifically what they were intended to measure. Some kinds of surveys are tested for validity on the basis of whether they correctly predict a future eventan election or some other test of btkhavioi, such as what proportion of the public plans to buy a certain product. Yet it cannot be determined whether the survey proportions were correct and remained constant until the election or purchase; whether they were incorrect, but the proportions changed and happened to coincide with the election or purchase when those occurred, or whether pLtblication of the results turned them into a self-fulfilling prophecy.. Stylies by survey- specialists Hugh J. Parry and Heln M Crossley have shown the validity of survey findings to be somewhat disappointing They showed that almost 25 percent of a sample of the electorate claimed they had voted in the 1944 elections wheny had not. Other studies show that more people tend to "remember" }laving voted for the winning candidate than actually did so. One method developed to validate survey data is to compare responses with such known dat.I as age or sex. Theoretically, if the ,

~~113 SURVEY RESEARCH AND OPINION POLLS~~

~~1 1 j J 4~~

sample i my random, all characteristics Of the populationincluding age, sex, education, and whatewr opinions may be held by the populationwould be reflect the sample in the same proportions as they exist infihe population as a whole. Thus; if the verifiable w characteristics are correct, everything else is consideredcorrecta leap of faith. Another Validating method involvescorrelating the index or. measure used with some other accepted measure. An index of religiosity might be the number of timeda person goes to church or it might be a minister's judgment. If the numberof times someone goes to church happens to coincide with the minister's judgmentof her or his religiosity, then the index h'as "criterion validity.:'A third validation method of survey research is knownas construct validity. Conserva-

tism, for example, is a construct thatmay be measured by a number of. 4 indices, including social and economic leveland politidal persuasion. L.- Survey findings regarding the conservatism ofa population have con-- struct validity if indices of conservatismare included and measured by the survey. Validity,is related to reliability, but it isnot the same. A watch may be highly reliable, meaning that it is consistent and doesnot lose or gain any lime. But it may not be valid: that is, it tells thewrong time; for example Paris time father than New Yorktime. There A no doubt that survey research can define, tigtilbe,and capsulize the prevailing attitudes, behavior, and characteristicsof a population. Without such techniques it wouldnot be possible to make group comparisons br predictions or to developnorms (standards against which to evaluate individualsor groups). But it must be remembered ithat estimates are whatare involvedalbeit the best estimates available... It has been estimated thatsome 2,000 survey research organizations conduct surveys, and thereare many times that number of businesses that conduct theirown studies. The technology is decep- tively easy to learn, and the rewardsare high. The United States i Congress often requires that atleasta part of any money that is appro- priated by the executive branch for projects of alltypes be set aside for evaluationfrequentlyisurvey research is used for thispurpose. Tire American Statistical Association has beenconcerned about this requirement and has sponsoreda study to assess survey practices. Of 26 federally sponsored survey's, 15 didnot meet their objectives because of pow design or sampling methods; technicalflaws, such as high refusal rates, not-at-homes,etc.; or poor supervision of inter- viewersor of coding and puhching of data. Seven of the 10 nonfederal

~~114 ONLY ONE SCIENCE t~~

1 14:. I,) surveys also failed to meet ASA objectives. The social science community is concerned ab9ut a variety of problems affecting, surveys. One of them is the growing number of unqualified firms and individuals who are conducting them. Another is the fact that the mass media occasionally either publish survey results without giving adequate information about the method of data collection or draw inappropriate conclusions from the results. A third is that unscrupulous salespeople occasionally insinuate themselves into a home on the pretext of conducting a survey and then switch to a sales pitch. Serious deficiencies remain in the scientific quality of some survey information, in the practices of the indUstry, in an over-reliance on surveys, and in the demands it places on a heretofore cooperative public. * A review of the positive and negative contributions of survey researcho American society suggests that it is, and h4s been, impor- tadas a planning resource, a managerial aid, a citizen information base, and a means of constructing and testing social theories. Like most other sophisticated tools in use'today, survey research must'becon- tinuously refined if it is to keep pace with the society it was designed to serve.

~~fit~~

~~t 115 SURVEY RESEARCH AND OPINIONOLLS~~

~~12i Iv~~

~~WO.~~

~~It-~~

~~116 ONLY ONE SCIENCE 'F 4 ) 424, '4 1:~~

~~VAT Pesticides and VA VA FAPest Control VA AT YAM~~

Since the first cave dweller took a swing at a flying gnat, pests J. have been a problem. Some are merely nuisances; otherspose serious threats to crops, forest products, livestock, and human health. For thousands of years effOrts to manage agricultural productivity were, at best, only moderately effective and resulted in relatively low and unpredictable production levels. Before the twentieth century most successes in achieving higherinricultural yields had been piece- meal. Farmers failedto consider or were unable to defeat the entire range of crop-limiting factors at once: If a farmer used an improved crotivariety, for instill-ice, the crop was 'otten inadequately fertilized or watered or was destroyed by pests. For years the only contrelk that kept plant and animal pests from completely overwh$1ming the ivorld's crops were naturalones genetic resistance to pests, weather adverse to pests, natural predators, and parasites. Plant and animal pests courd, and sometimes did, destroy entire harvests. The Bible recounts plagyes of locusts that "covered the face of the... earth" and ate "eve.v.-herb of the land and all the fruit of the trees." In the1840s in Ireland the late-blight of potatoes (Phytophihora infesity2s) brOught about the deaths ofmore than 750,000 10 people and compelled the emigration,ofat' least a million more. More recently, grasshopper hordes decimated yegetation in the prairiesof 4,1\1orth America,leaving behind a land stripped bare. These "plagues"

Because helicopters can spray were, fortunately, the exception; but when they occurred, they, were. , smallaimd irregularly-shaped' devastating. arsiW Wye easily than a small aircraft can, they may be used It was not until about 40 yelars ago that scientists working in f. apply Fij ,i,iesVII, heti chemistry, plant genetics, entomology, engineering, and a host of other coptrois treating an orange" _grove near Lyke Wales, Florida fields began to develop truly effective methods for improving and with a funp,ide ensuring crop yields. One of the most dramatic successes was in giving the farmer new.weapons emical pesticidesto use in the age-

~~117 PESTICIDES AND PEST CONTROL 4 -~~

112 old annual battle against those organisms that threatened theircrops and livestock. The term "pe,st'' means different things to different people. To many, it means only harmful insects and rodents; but the term also includes noxious organisms, such as mites, nematodes, weeds, andcer- tain vertebrates. In addition, essentially all organiSms that cause pla'nt disease, whether fungi, bacteria, nematodes, or viruses,are considered pests. Still, hpre is a contirrIting semantic problem about what is and what is nkg.pest. Butterflies are beautiful, but their larvae may be serious plagtpests. The wheeling flight of a flock of starlingsmay be a

beaUtiful sight to a poet or a city dweller, but toa farmer with a. "1/4 newly-seeded field it is a nightmare. Birds are also considered pests when they destroy crops such as cherries and strawberries. Clearly, honeybees are normally highly desirable insects, but theymay be pests to individuals who are particularly sensitive to their sting or to others who find themselves unwittingly surrounded by a swarm of bees. Circumstances of quantity, location, timing, and culture also affect the status of an organism as a pest. Bermuda grassmay be useful erosion control and as a source of hay, but it can be a damaging weed in agriculture or a problem in home gardens. Dogs are prizedas pets, but stray dogs may be legitimately classified as pests. Most pest control measures, however, are directed at species that are generally recognized as pests. The control achieved can be con- enefit to society, whether it results in reduction in-the trap ion of malaria or plague or in the number of grasshoppers; higher yields of sweet corn, wheat, or apples; or less damage toroses in a garden. While the exact amount of world-wide crop loss attributable to pests is unknown, it is estimated to be about one-third of the potential food productionfood for about 1 billion people. Rice has the greatest loss from pestspossibly ashigh as 40 to 50 percent of potentialpro- duction. . The magnitude of the pest problem can be appreciatedeven more when it is recognized that the amount of land now under cultivation in the United States is half again as much as would be required if there were no pest-induced losses. The crop-related energy needs.and amounts of labor, capital, fertilizer, and other requirements alsoare correspondingly greater. The total costs are huge. The United States, particularly, has seen a shift from small farms that grew a wide range of crops to large, specialigedfirms that engage

~~(14%~~

~~118 ONLY ONE SCIENCE~~

4 in monoculturethe intensive production of single crops. This development has provided pests with4 highly concentrated food source, in ar1/41 around which they can breed and feed with unpre- ''" cedented ease. Thtie-is little doubt that monoculture has led to ,a greater need for pest control:ithout monoculture, however, the population would have been more permanently bound to the land, and technological development probably would not have occurred as rapidly as it has.. t

A Civilization must defend itself not only against organisms that 0 destroy buildings, clothing, crops, and stored food but also againstcer- tain pests that adversely affect health. If epidemic typhus, malaria, trypanosomiasi ncephalitis, and other diseases that have caused widespread d ath and debilitation are to be avoided, defense against disease-tqn mitting creatures must be carried out continuously. Pests that bothei people and higher anipary buzzing, crawling, stinging, biting, or sucking blood must also be controlled.

This photograph, taken in1946, shows DDT being applied kill corn borers in an El Paso, Illinois cornfield The photograph also illustrates the agricultural practice of monoculture USDA Use of the pesticide DDT (dichlorodiphenyltrithloroethane) brought about a dramatic change in the world. It was less than two generations ago that victims of typhus and malaria were dying by the 4 thousands in the United States and by the millions aroind the world. By helping to control the spread of insect-borne diseases, DDT saved millions of lives and prevented millions of people from becoming sick during the early years of its use.

~~119 PESTICIDES AND PEST C0141110t-~~

~~1 2 5 't~~

\ ithOut the development ofDDT and its continueduse against the pests which carry typhus and malaria(body lice, rat fleas, ticks, and Anopheles mosquitoes) it is likely thatthe ravages of these two ki er diseases would still be a major factor in kortalityfor the United States Qs for other countries where they have been essentiallyeliminated. DDT and other Pesticides also have beenused to control the spread of yellow fever, bubonic plague, certain typesof encephalitis, and a host of other diseases:

= In the early INOs, Americans became ch more aware of the uses and the dangers of pesticide. Theygreespecially conscious that some of thele toxic chemicals persist in thee vironment and accumulate,in certain plants and.animalsand even in human beings. Rachel Carson's book, Silent Spring, spearheadedtheir interest. A report prepared by President Kennedy's ScienceAdvisory Committee in1963 also contributed to thisawareness and called for a reduction of theuse of chemical pesticides. These publicationshelped set the stage for rapid expansion of public involvement andconcern. Since that time both the scientific comrnunity4ind the publichave become increasinglyaware of the need to weigh the risks of thistechnology against its benefits. A substantial effort has been made to rimirtheuse of persistent pesticides and to replace them with ecologicallysafer chemicals or other methods of pest control. Alternativemethods are also being sought because, over time, pests frequentlxdevelop tolerance, resistance,or immunity to particular chemicalsoriontrol measures. Cultural methods (crop rotation, tillage, propa.timing of planting, and fertilizer and water management)are now being used more extensively. Other approaches to pest control include biologicalcontrol (regulating pests through their natural enemies, includingvertebrates, insects, mites, tingi, bacteria, and viruses); theuse of behaviorynod4ing substances (pheromones), sterilization techniques,eradication programs, and the use of synthetic hormones to disrupt critical stages inlife cycles. Current and potential problems withchemical pesticides demon- St-rate the need for increased researchto develop more and better alternative techniques. Until those techniquesare developed, however, ti s.ome pest problems exist for which chemical controlsare the only aliailable answer.

-13ecause.of the adaptability ofpests, possible hazards to the environment, the economic impact ofpest combinations, andconcerns -about the use of certain chemical pesticides,there is great interest in developing and refining'an approachto pest control which would promise bbth more stable crop productionand protection and the least

~~12b ONLY ONE SCIENCE~~

12 possible risk to human beings and to the environment. That approach . is called Integrated Pest Management (IPM). 4 The term, "Integrated Pest Managemerit," has been-defined ina variety of ways. It is described in a 1979 repOrt of the Office of Tech- nologyAssessment (OTA) of the Congress of the United Statisas "an all-inclusive concept that should be applicable to all pests (weeds, plant pathogens, nematodes, vertebrates,in.sec4s, etc.). However,terminology, control tactics, and strategiesvary among disciplinary groups, so that it is difficult to arrive at a definition completely appropriate to all interests." In that report, OTA uses the definition that IPM "is the optimization Of pest control inan economically and

ecologically sound manner, accomplished by the coordinateduse of . multiple tactics (Tactics are the specific methods used to achievepest control' These include chemical pesticides, pest-resistant varieties, cultural practices, biological control, and others.) toassure stable crop production and to maintain pest damage below the economic injury level, while minimizing hazards to human beings, animals, plants and the environment." Kkey to understanding the role of pesticidesin society today, both in regard to the benefits of increased crop yields and improved public health, is to recognize that pesticides have not eliminated the problems of pest control. They have allowed society partiallyto control the damaging effects. If pest managementwere to stop, the life- threatening situations and crop and livestock destruction caused by pests would rapidly reappear. A HISTORICAL VIEW OF PESTICIDE ANAN DEVELOPMENT The practice of applying various substances to plants, soil, livestock, and sometimes even to people for the control ofpests goes back many years. Fine dust or ashes sprinkled on leaves of crops to combat ) leaf-eating worms-were among the earliest pesticides. Saltwas occasionally used to discourage the growth of weedsamong certain salt- tolerant crops. A story about how South American natives killer fish by using the root of a certain plant (Derris elliptica) ledto the discovery of rotenone, an insecticidal compound. This substancgwas applied as a dust long before the science of chemistry perrriitted identification of the active complex chemical ingredients of this natural product. The . insecticidal properties of a powder derived from dried flowers of the ' daisy-like Chrysanthemum cocczneum plant were discovered in the early

~~121 PESTICIDES AND PEST CONTROL~~

- ,,. I' 1800s when it was obierved that insects avoided this plant. Since then, the specific chemicals, pyretlicins, which gave the flowers their insect-killing power, have beenextracted and have been synthesized. ../ . Farmers used arsenicals such as Paris "grien as early as 1865 to combat insects which were damaging potatoes in the Rocky Mountain region. Fumigatiim with hydrogen cyanide and lime sulfur washes were used in the 1880sto fight insect pests in California. Although these inorganic compounds were better than nothing, they had to'be used in high doses and even then were sometimes notvery effective. They.were alsohighly toxic to human beings. . Legend has it that the develop6ent of one chemical pesticidewas brought about by a group of thirsty French thieves. To stop theirnoc- turnal raids ipt6 the vineyards of Bordeaux to stealgrapes, vintners developed I combination of copper sulfate and lithe. Thegrapes were sprayed with this "Bordeaux Mixture" to keep the "human pests" away. The mixture left an unappealing blue residue, copper hydroxide, and made the grapes toxic if they werenot thoroughly washed. The human pesticide worked. One day an alert grower noticed that the mixture also discouraged growth of mildew, so after the thieves were no longer a problem the vintners continued their sprayin&and a . . chemical pesfidde was,born. '-::1-- .... In 1919 E. B. Blakeslee found that paradichlorobenzene could be used to protect clothes against moths; later it was found that thissame, chemical was also effective against peachtree borers (Synathedon exitiosa).

~~Horse-drawn power dusters were used in the early 1900s In the operation depicted, sitlphur or arsenate of lead is being dusted on trees as thtwagon is slowly pulled down the road USDA SEA~~

~~122 ONLY ONE SCIENCE . .,,~~

/ 1 t I.. 1 28 1 At about the same time, it was discovered thatsome chemical compounds had valuable insecticidalas well as fungicidal properties, and systematic searches for such laboratory-produced pesticideswere undertaken in earnest. The early basic research on pest-control techniqueswas done mainly by biologists who concentratedon classifying particular types of pests and studying their life cycles, engineersconcerned with the control of mosquitoes by drainingswamps, and chemists who worked to improve, the few chemicals known to be pesticidal. In the earlypart of the twentieth century, scientists/Were searchingspecifically for parasites and predators of importaint pests. Biologicalcontrol was almost all they could offer because therewere few effective chemical pesticides. Eventually chemists beCame involvedon a more practical level. They improved formulations, refined theproducts, and identified chemical structures. Through research theyexplored methods of synthesizing chemical pesticides and of improving waysloapply them Closely related to this latter researchwas the start of research on the behavior of emulsions, spray droplets, dusts, andother physical states important to the application of pesticides.Studies such as these involved physics and engineeringas vegll as chemistry and biology. Basic studies of kinetics of charged particles and emulsioristruc- ture in the 1930s allowed scientists to prepare chemicals that would permit formation of an emulsion suitably stable for long-termstorage. Emulsions were needed in which chemicals remained uniformly suspended when diluted in the spray tank before application,but in which the-droplets of the emulsion would break immediatelyupon contact with the sprayed surface so the active pesticideROuld adhere. Formulations and equipment for applying pesticideswere modified'to increase-the concentratioraf the active ingredient in thespray tank, thus lowering th?volume of diluent which had to be transportedand applied. The basic patent for the pressurized gas aerosol method of applying substances was granted to two chemists in the UnitedStates Department of AgricultureLyle Goodhue and W. Sullivan, Aerosols later became popular for many household'uses. In 1874 a German chemist, Othmar Zeidler, synthesizedDDT and filed it away as a curiosity. In 1939 that curiosity provided pesticides' quantum jump when another chemist, Paul Muller, at .1. R. Geigy, S. A., in Switzerland, discovered theuse of DDT as a pesticide (for which he received the Nobel Prize in Medicine and Physiologyin 1948). Besides being credited with saving millions of lives which might

~~123 PESTICIDES AND PEST CONTROL O~~

have been lost to insect-borne typhus and malaria%during and after World War H, the discovery of DDT set off a new effort of basic research into the chemical control of pests. This new effort was led by the establishment of the United States Office of Scientific Research and Development, a wartime group of outstanding scientists from many disciplines. Among the responsibilities with which this group was charged was that of finding new and improved methods of controlling typhus, malaria:and other infectious diseases. This challenge required the cooperation of organic chemists to develop new compounds, physiologists to determine the mode of action for pesticidal chemicals, sand toxicologists to determine hazards involved in the use of the new chemicals. .1,- By the end of World War II, this team effort proved so successful it was rapidly expanded to apply to agricultural pests, including weeds. Basic studies on the interaction between chemical' structure and biological activity led to discOvery'of such new Torms of pest control as the hormone-like activity of 2,4-dichlorophenoxyacetic acid (1,4-D) on plants, natural and synthetic insect hormones, and pheromones. . For a number of years after World War II, DDT and other'chlorinated hydrocarbons were dominant in the insecticide market. Over time it became apparent that such stable substances can be dangerous because they accumulate and migrate in the environment. Moreover, pesticide resistant mutations, which sometimeyseemed to cause even more destruction-than their ancestors, were beginning to emerge.

' Resistant strains result from selective survival of individuals hav- fima genetic constitution which provides immunity. Although such resistance has been described since the early 1900s, it began to increase in the middleof the 1940s when.large scale application of pesticides n became common practice. In some cases insects have developed resistance to as many as foUr different toxic chemicals or to other pest con- trOl measures. Alternative pesticides which did not pose threats to the environment were sought. As a result, organopho'sphorus insecticides (such as malathione and parathione) a/id carbamate insecticides (such as Sevin® and Furadan®) came into increasing use. The organophosphates were' developed from a research base'that began in Germany in the early 1930s at I. G. Farbeninclustrie. Sane of these cherriicals are systemic (absorbed by a plant through its leaves, stems, or roots) causing a treated plant to become lethal for a time to insects that eat it. (Nonsystemic insicides are lethal to insects that touch or ingest them directly) gafiophosphates can be used to control disease-

124 ONLY ONE SCIENCE carrying mosquitoes and animal pests suchas fleas, ticks, cattle grubs, and hornflies The organophosphates remain in theenvironment for a shorter time thah the chlorinated hydrocarbons and, therefore,have limited uses. Despite this relative lack of persistance,some organophosphates have been found tribe dangerously toxicto human beings. Development of the carbamate insecticides began in the1940s at J. R Geigy, S. A., in Switzerland. Most of these substancesact topi- .4 cally, but a few show a high degree of systemic activity.Although they do not tend to leave harmful depositson foods, some are harmful to warmlooded animals. Other carbamatesare effective because they act onbiologicalreactions and processes that are common,in insects but are not as common'in mammals. Recently, the evaluation of tly effects of pesticides hasbeen ,greatly influenced by the tremendous acianci,sin sensitivity and specificity of chemical analytical technology and techniques.Before the 1950s only relatively high concentrations of pesticidesand their metabolites could be identified and measured. Residues ofless than .1.,/one,part per million of pesticides in animal or plant tissueswere considered analytically negligible because theywere regarded as trace residues that could not be measured. In recent years the develOpment of improved analyticaltech- niques, however, has provided scientists with the abilityto analyze and identify certain pesticides and other chemicalsin concentrations as low as one part per trillion, (One part ina trillion is the equivalent of about 1 minu/in 1,900,000 years.) People have becomemore concerned about pesticides at these lower concentrations, andsome research has shown that, even at low concentrations,significant hfan risks can result from the persistence of some chemicals in the e irbnment.Although the defenders of pesticides characterize such minute concentrations as "one drop of vermouth ina tank car of gin," others take the position that even these low concentrationsof certain substances are unacceptable Another effect of the driveto achieve pesticide use that is almost or completely risk-free hasbeen to intensify the interest in Integrated Pest Management. 4 THE DEBATE OVER CHEMICAL PESTICIDES Just as any technology may yieldnew benefitsreal or imaginedit can also producenew hazardsalso real or imagined. A conflict over which hazards and benefitsare real and which are imagined seems inevitable.,

~~125 PESTICIDES AND PEST CONTROL~~

~~131 /~~

. . Pest control by chemicals is a relatively new technology which most people agree should be integrated with biological, cultural, genetic, and other means of pest control. But any evaluation of the use of pesticides must also include an analysis of what happens when no 4 N. pesticides are used. Obviously, both sides of the pesticide ledger must be totaled before decisions on the greatest net benefit can be made. Evaluating the trade-offs in pest control methodg is often a matter of dealing with a moving target. Circumstances *changing contin- ually. New regulations and technical improvements and discoveries that reduce risks must be part of one side of the equation, while known or potential hazards that are detected in initial safety screening tests .. become a part of the other side. , Many agriculturists believe that those who are concerned with r health must balance the productiOn ledger against the possibility that an occasional person might be injured now or in the future as a result of trace pesticide residues. Those in agribusiness point out that the use, of nonchemical methods of pest management alone would result ina severe drop in crop production. Manyfear that the public health may be more adversely affected by reduced crop yields and increased incidence of pest-borne diseas-es than by the occurrence ofcancers, birth defects, or other illnesses in a few individuals. The potential of some pesticides to cause human illness and even death has been recognized for some time by government regulators and the agricultural chemical companies. D' ections for use and label precautions in the United States are eSsentiy legal documents, deviation from them is a violation of the law. The time that must elapse fol- . lowing application before harvest or slaugh er, the interval required before workers can re-enter a sprayed fieland safe disposal feature for used containers are-all a part of the pc ionary statements. In addition, pesticide products are classified into restricted and nonre- stricted groups; only a certified farmer or commercial applicator can, use restricted materials. And some pesticides-can cause problems even,

when used as directed. - . On the assumption that human beings must protect themselves, their crops; their,forests, and their livestock from pests by one means or another, all available information must be evaluated auil understood' before decisions can be made wisely. Those who began tAir gardening before the end of W orld War II will recall the skull-and-crossbones sticker required on packages of arsenical insecticides. Although some people were concerned about residues of these inorganic-substances on apples and other fruits, pesticide regulation was minimal. When the

~~126 ONLY ONE SCIENCE Y r~~

~~1-'32 I 1~~

~~This trap is used to monitor the number of codltrig moths in an area winkee sterile males have been released to control the population of this pest APHISUSDA~~

codling moth (Laspeyresia pomonella) developed resistance to lead arsenate, gardeners increased the number of applications to combat the pest- with the result that even drastic washings would not remove enough of the pesticide to meet the established tolerance levels. Then, basedon a' study, done by the United States Public Health Service, the Food and. Drug Administration raised the tolerance levels for this insecticide. With the introduction of a wide range of pest control materialssoon after 1945, however, product registration procedures with requirements for safety and efficacy 'testing prior to marketing were established. A statement of recommended handling piecautions, as wellas use directions, was required on the label.

Federal Regulation of During the earlr1950s people became more concerned that Pesticides exposure to certain substances such as cigarette smoke and asbestos might increase the probability of developing cancer. One,,consequence of this concern was that in 1958 Congress included in the Federal Food, Drug, and Cosmetic Act what is now known as the Delaney Clause. What clause prohibits the use of any food additives that are found to be carcinogenic to humanbeings or to animals. The ruling applies only to substances that are added purposely to food; it does not apply to

~~127 PESTICIDES,AND PEST CONTROL~~

~~133 I. v~~

,t. .. natural ingredients of foods, inadvertant or unavoic}6,ble contam- inants, or to pesticides applied to crops in the field. It doescover pesticides-that might coptaminate food as the result of pest control tech- niqu'es during storage''r processing of food as well as resies of pesticides if they become more concentrated during storageor pressing of the food. The clause neither offers guidelines as to what testsor carcinogenicity are "appropriate' nor requires testing for carcinogenicity,

although the requirement is implied. , .. In 1964 the UnitaFI States Department of Agriculture decided to refuse to register pesticides foi use on food crops unless the Food and Drug Administration established a finite tolerance fora residue of that pesticide on a particular crop or food. This action prohibited the registration of potentially carcinogenic pesticides for suchuses. Some experts thought that carcinogenicity testing should be required for all newly registered pesticides. At about thesame time, representatives o the agricultural chemiCal industry urged standardization of testing required for registration of pesticides so they could complete testing without undue delay before seeking registration. Inresponse to the request, in the early 1970s the Environmental Protection Agency (E attempted to develop complete guidelines for testingnecessary to, register a pesticide. Many tests that could,be used for one pesticide, however, could not be used for another; the task was abandoned:. EPA then proposed that complete carcinogenicity testing be required for those pesticides expected to result in residues on food, foodcrops, or feed crops; on those substances suspected of being carcinogenic because of chemical or other similarities to known carcinogens; and for those substances that tested positively in the Ames Test for mutageni- city. (The Ames Test is used to determine whether or nota substance is likely to cause inheritable mutations.) .. Prior to the Delaney clause, premarketing safety tests for new pesticides did not place major emphasis on testing for Carcinogenicity. Subsequently, a wide-range retesting (with laboratory animals) of pesticides' already on the market was instituted. Traditionally, in designing long-term feeding tests, toxicologists have used at leastone concentration of the substances to be tested that is high enough to induce some effect in laboratory animals. Because many substances being investigated show no effect at doses comparable with those to which human beings are EyEically exposed, it is sometimesnecessary to use relatively high dosales to obtain any effect. The proper interpreta ion of data about o-arcinogenicity from these high-dosage-level tests cotitutes one of the most frequently debated aspects of pesticide contro ersy. One group of people claims that even

~~T. 128 ONLY ONE CIENCE * 0~~

r 1 a single molecule of a carcinogen can, over time, produce cancer if it damages the genetic material of a cell. They theorize that, if tumors in laboratory animals can be produced in tests usingsrossly exaggerated dosages or using very large numbers of animals At low dosesover a long period of time, human cancers can result from infinitely 4naller,, doses. When this belief is added to the ability of chemiststo detect toxic substances at concentrations as low as one part in a trillion, to some people the world seems to become a carcinogenic nightmare. Otherpeople are convinced that there are concentrations ofmost toxic substarites below which no harmful effects willoccur. Neither group can convince the other. ,' Where reliable epidemiological data exist, theyare useful in deter- mining carcinoginicitY. Unfortunately, epidemiology is of limited value in most instances where chemical pesticides are concerned primarily because, for most chemical carcinogens, itmay take as long as 20 years or more between the first chemical exposure and the onset of cancer in human beings. By the timea substance is confirmed asA carcinogen, many people may already have been exposed to it. Epi- demiologists also have trouble locating people whomay have been exposed years previously to a particular substance. When'and if such individuals are found, they often have also been exposed to several other potentially hazardous agents. Thus it is difficult toprove that a pesticide is carcinogenic unless there is a wet-defined, easily traceable,

~~)J relatively homogeneous group of people whO have had similarexpo- sure to a particular substance.~~

In 1972 the Congress passed another piece of legislation which has. had a great effect on the use of pesticides in the United States. The Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) places the responsibility for proving the safety of a pesticide on those who wish 'to make themnot on the general public as represented by the EPA.

I. 2,4, 5-T In 1970 the majority of pesticides (an estimated 90 percent)were organic chemicals. Insecticides and fumigants, including nematocides and rodenticides, accounted for about 47 percent of the organics. The other organics were divided between herbicides (37 percent) and fungicides (16 percent). Between 1966 and 1976, the use of pesticides in the United States doubled to more than 600 million pounds. (Theuse of , insecticides and fungicides increased only slightly during this period.) The model represents the chemical structure of 2, 4, 5-trichloro- The much greater use of herbicides is responsible for most of this phenolya,etu and (2, 4,...5-T) growth. In 1977 herbicides accounted for 58 percent of all pesticide u; ,,,L II, witatc.:,, v.,an.,,, sales in.the United States. Although the total use of chemical pesticides

129 PESTICIDES AND PEST CONTROL , ( 135 in the United States is not expected to increase appreciablyover the next 15 years, the use of herbicides is expected to accelerateconsider- ably. The herbicide 2,4,5-T (2,4,5-trichlorophenoxyaceticacid) has been in us for abou. t 30 years in the United States, Australia, NewZealand, Germ'ny,and a number of other countries. Eachyear about 3.8 million acres in the United.ates are treated with 2,4,5-Tprincipallyto control brus and weeds On land used for right-of-ways, timber,rice, and grazing. her methods for controlling these types of vegetationgen- erally are n t as effective or economical 2,4,5 -T. For example, if the In this photo,2, 4, 5-T ts being sprayed to control brush and current use o 2,4,5-T on United Staffs land were cancelled and Lqi lanst land This herbi- other known s viculture methodswere used, it is estimated that the (ide has also been sprayed along foreSt industrr right-of-ways, on timber, and ould sustain a cumulative net income loss of about rke-gruwing areas to ,ontrol $800 million over 0 years. If 2,4,5-T could not be usedon right-of- the growth of weeds ways, it is estimated that expenses for right-of-way maintenance Dt,'OIVI,IlltaCOrttrat11),1 would increase 35 percent (about $35 million) annually"' S.

(

~~130 ONLY ONE SCIENCE~~

13" IX* IF The dramatic contrcwersy wl-pch surrounds the use of the herbicide 2,4,5-T is One example of some of the complex scientific; political, and economic considerations which canbe associated with the use of certain pest-control methods. Late in the 1960s itwas reported that high dosages of 2,4,5-T caused birth defects in experimental animals. In fact, it Was not the herbicide itself that produced the birth defe9tsifi

..these animals, but rather a highly toxic impuritythe dioxin TCDD . (2,3,7,8-tetrachlorodibenzo=p-dioxin) (shown on page 129) that occurs in the fdenulation of 2,4,5-T The presence of TCDD is the principal reason for theconcerns about the herbicide. The herbicide 2,4,5-T was a key component of "agent orange," a defoliant widely used ifi jungles during the war/in Vietnam. The discovery thaf2,4,5-T had the potential of causing birth defects was instrumNal in discontin g the use of "agent orange" in Vietnam. Currently a number of eterans of that war allege that their exposure to "agent orange" h eriously damaged their health by increasing their rClc of cancer ancrcausin&them genetic damage. The validity of these rges is yet to be deterined. Prior to the late 1960s certain quantities of 2,4,5-T contained up to 30 parts per million (ppm) of TCDD. When the Dow Ghemical Company, the principal manufacturer, learned of the high toXicity_bf TCDD, it changed its manufactur- .. ingprocesses, so that the herbicide produced by 1969 lipped less . than 1 ppm. In 1970, not long after the ban of "agent orange " for military uses, the Unfrd States Department of Agriculture, United States, Depart- ment or the Interior, and the United State 1 ep rtthent Of Health, Education, and Welfare jointly agreed f4rohi t.certainuses of

~~2,4,-T. The Environmental Protection Agency (EPA) set allearing 4 date in 1974 on cancellation of the remaining registered uses, of 2,4,5-T.~~

It }fad been found that TCDD disintegrates rapidly (that is, within days) in some environments receiving bright sunlight, thereby theoret-- ically reduling the risk of using 2,4,5-T on lie() fields, on right -of- ways,and on rangelands or pastures. On the other hand, it was not known how rapidly TCDD breaks down under certain other conditions, such as after being sprayed on forests. Moreover, it was not clear whether or not TCDD would concentrate in animal tissues, as do 'Some other chlorinated hS?drocarboivs. Furthermore, only a handful of la,boratqries in the world had the highly trained technicians and sophisticated equipment necessary to detect very low concentrations,.

~~of the chemical. #. By thelime scheduled for the public hearings, the most advanced i '14 Oft~~

~~131 PESTICIDES AND PEST CONTROL~~

a methods ofirnass spectroscopy had so increased the sensitivity of aoalysifr that as little as 10 partsper trillion (ppt) of TCDD could be detected in some tissues, but the method had not been verifiedin a sufficient number of laboratories to be useful in the hearing.Without infor on abotit human exposure (usually derived primarily through analysisdiet residues, tissues, of milk) there could beno sound evidence of uch exposure or, therefore,-of risk to human health.In spite of scatred reports of problems in human beings and animalsat-' tributed.. to 2,4,5-Lthe ub is hearings scheduled for1974 were cancelled because of a lack of sufficient data., In the meantime, the EPA created a processitlled the "rebuttable presumption against registration" (RPAR) to help identify pesticide uses-which may not satisfy the legal reciuirementsfor a pesticide's registration. The procedure also establishes a method for gathering and evaluating infoimation about the risks and benefits of these A uses and allows for public participation in theprocess. In effect, for a pesticide to be registered for a use, its benefits must exceed its risks for thatuse. A The RPAR for all pesticides containing 2,4,5-T productswas issued in 1978. During this process, in March of 197.9, the EPA announcedan emergency suspension of spraying of 2,4,5-T on forests, in pastures, and along right-of-ways. Use of 2,4,5-Ton rangeland andrice were not included in this suspension order. The EPA Saidit did not say "that the health effects in humans are positively proven,or that 2,4,5-T should never' be used again." But it did say that "there is sufficient evidence to stop further exposure of the chemical until the issuescan be resolved." The evidence (collected during the RPAR review) to which theywere referring was, in large part, information derived from the results of animal studies and an epidemiological study. The principal impetus for theemergency suspension of the use of

A 2,4,5-T was an epidemiological study said to showa "high probability that the herbicide is linked to human miscarriages-fnan area where 2,4,5-T is used regularly." The study (called Alsea II because thiswas the second study conducted in the Alsea Basin region in Oregon) which precipitated the suspension has since been severely criticized by a numbei of independent experts. Their criticisms include incomplete and inaccurate data on 2,4,5-T usage, inaccurate dataon spontaneous " abortions and failure to recognize that the monthly variatiorlin rates of hospitalizedspontaneous abortions described in this studywas no greater than might be expected through random variations.. Thecon- sensus of those who reviewed the Alsea II study (includinga number of scient. is at a 1979 conference on 2,4,5-T sponsored by theAmeri-, carjrlureau Federation) was that it did notseem to warrant the

~~132 ONLY ONE SCIENCE 00~~

13s EPA's conclusions. On the other hand, if has been shown in various studies of certain species of test animals that TCDDcan cause cleft palate, kidney abnortnalities, fetal mortality, and reduced fetal weight. Even though sophisticated techniques were used in these studies, traces of TCDD were found in a few cattle in some areas where 2,4,5-T had been used, while in a number of other studies no traces of TCDD have been detected in animals. Some scientists maintain that more sensitive instruments might have detected TCDD and that even extremely low concentrations arehazardous. Furthermore, people do not yet agree on all aspectsof interpreting the dose levelresponses, in other words, whether or not there is a level of this chemical below which no effect will occur. The question algo remains whether or not 2,4,5-T and/or its contaminant TCDD might pose a threat of cancer, birth defects, or abortion to human populations. Certain groups vehe- mently continue to oppose any use of 2,4,5-T no matter how little. TCDD it contains. Currently, the re fommended aximum level of TCDD in 2,4,5-T is 0.10 ppm. Dow heroical Com ny reports that more recent technology has made it possible to r ce the level of TCDD in the 2,4,5-T it produces to less than 0.02 ppm: When the use of a substance such as 2,4,5-T is suspended, hearings can be requested on the cancellation of that chemical's registration. In July of 1979 the EPA issued notice of an intent to hold a hearing to determine whether or not the remaining nonsuspended registered uses of the pesticide (on rangeland and rice, and for noncrop uses) should be cancelled. In September of 1979 the Federal Insecticide, Fungicide, and Rodenticide Act Scientific Advisory Panel recom= mended further reductions of the level of TCDD in 2,4,5-T, obtaining more long-range data on exposure, and evaluation of information from ...,...animal studies that had not yet been completed. The Scientific Advisory Panel also said it "could find no evidence of an immediate or substantial hazard to human health or to the environment associated

with the use of 2,4,5-T" on rice, rangeland, orchards, sugarcane, and in... certain noncrop uses. It therefore recommended that the EPA not hold- a meeting at that time to resolve the fate of the nonsuspended uses.

Nevei'theless, the EPAcurrently conducting hearings on both , the suspended and nonsuspended uses of 2,4,5-T. Evidence is being collected from the results of animal and epidemiological studies, and new facts are emerging. For examples it has been reported that many of the 75 chldrinated dioxins may be generated from sources other than 2,4,5-T, such as fuels and other common materials. The full significance of this finding is still not understood, but when an environment )ep'osed to 2,4,5-T is examined for TCDD, measurable amounts of

~~133 PESTICIDES AND PEST CONTROL~~

~~1.3J I v~~

TCDD may or may not be found. However, when there isa search for TCDD and other chlorodioxins in an environment exposed to large sources of combustion by-products, measurable amounts of TCDDare found. More information is being assembled as various studiesare completed. For example, scientists were recently unable to detect, dioxin in mother's milk in areas where 2,4,5-T had been-used. No residues were detected 40g the most modern scientific equipment equipment capable of measuring residues of 1 to 4 Ppt. It is not known whether any dioxin is present below this limit of detection. In December of 1980 the Advisory Commission on Pesticides forthe United Kingdom issued its most current report concerning the safety of 2,4,5-T. That report concluded "that 2,4,5-T herbicidescan safely be used-in the UnitedUnite ingdom in the recommended way and for the recommended turpo ; however, the committee also -expressed its intention to "continue td examine any soundly based new evidence or

information." . ,, The hearings on the 2,4,5-T issue in the United States are still\ underway. Becadse of.the many complex issues involved, it probably will be some time before they are concluded and important risk/benefit questions about 2,4,5-T and a host of other chemical. t. pesticidei are resolved.

ALTERNATIVE METHODS OF PEST CONTROL he publie0 mobilized by the publication of Silent Spring and the reportI-1President Kennedy's Sdence Advisory CoMmittee has broughtal:rut a serious search for alternatives to the heavyuse of chemat pesticides. To achieve the goal of reducing the use of chemical pesticides, the concept of Integrated Pest Management has become a research priority for i9i.d ries, universities, and government. A number Of new, nonchemi methods of pest control have come understudy and have prdgressed to field testing with varying degrees ofsuccess.

Biological Controls- tl-he greatest efforts in biological control have been directed:at insect% and mites.,Hundreds of years ago the Chinese usedants to control insect pests on citrus plants. But it was not until around 1890 thatA major biological control effort was undertaken in the United States.

~~134' ONLY ONE SCIENCE A~~

At that time the cottony cushion scale, Icerya purchasi, nearly,destroyed the beginnings of citrus farming in Calif ornia..This scale was found in Australia but was not considered a pest there because of a natural enemythe Predatory vedalia beetle, Rodolia cardinalis. Within a year after the vedalia beetle was imported and established in the Cali- fornia citrus groves, the scale problem was under control.

Around 190 theAtanycush- ion scale (right), threatened the success of citrus farming in CaliforniaIts mama' enemy, the predatory vedalia beetle (left) was used to control this pest in the first maior biological con- trot effort thtt was undertaken in the United States APHIS USDA The suppression of this pest continued for more than 60 years until DDT was used in some groves to combat other pests. Unfor- tunately the DDT killed the beetles, too, and there was a widespread recurrence of cottony cushion scale. Control was achieved through the reintroduction of the vedalia and by adjusting the spray treatments. Insect ciiitases can be important natural regulators of insect populations, but it is difficult to predict where and when an outbreak of a pathogen will occur. There have been some cases of successfully introducing an insect pathogen to control certain insect pests. One notable such program was designed to control the Japanese beetle (Popilha japon- ica), an insect that injures shade trees, ornamentals, lawns, and various food crops. Estimates of the average damage caused.each year in the United States by this beetle range up to $20 million. The beetle lays its eggs in the soil, and the eggs hatch into larvae that eat plant roots before they enter their inactive pupal stage. The adult beetles feed on a

~~135 'PESTICIDES AND PEST CONTROL~~

~~;14-e 4-~~

~~A %W.~~

several types of ripe fruits, especially peaches. It was found that a bacterium, Bacillus popilliae, can infect the beetle larvae. This pathogen is propagated by injecting the live larvae of the Japanese beetle. Later, the spores of the bacterium; are removed from the larvae, dried, and mixed with talc to achieve the desired concentration of spores. The "spore dust" is then spread on the soil to attack the larval beetles. The infected larvae further inoculate the soil and add to the concentration of spores to the point that the beetle population is controlled. In tie eastern United States, this method of control is one of the most important biological factors in regulating the number of Japanese beetles. Another commercially successful effort to control insects with bacteria has been the use of a bacterium named Bacillus Mums:ens:5 to control pests such as the cabbage looper,richoplusia ni, and the corn earworm, Het:OM:5 zea. B. thunngiens:s, h been used commercially for about ye ,but it does have its drawbacks. It and other microbial contrs are highly specific, chemical insecticides must, therefore, be used to control the aphids, mites, and other pes,ts in the same field. The bacteria do not persist in the field, so applications must be made frequently, as with conventional insecticides. Furthermore, bacterial insecticides do not produce the quick knockout of pests that farmers like to see. Another successful use of biological methods is in the control of the prickly pear cactus, Opu` ntia strula, on about 60 million acres in Aus- tralia. This unwanted plant has been controlled by introducing moths, Cactoblashs cactorum, which feed on it. Generally, howeVer, the biological control of a single species of weed is not practical because other weeds quickly take over when the weed causing the primary problem is gone. The successes of biological controls have been limited.Unfor-- tunately, of the approximately 85,000 species of insects living in the United States and Canada, only about 60 species are reported to have been partially, substantially, or completely controlled by managed biological intervention.

Sterile-Male Technique Sterilization of male insects by radiation as a mean; of pest control or eradication was conceived by E. F. Knipling some 40 years ago. His theories were first put into practice in 1955 to eradicate the screw- worm, Cochliomyia hommvorax, in Curacao. Screwworms are the maggots of blowflies. The maggots'eat the flesh of living cattle, and the wounds they cause may kill the cattle or severely damage the quality of the cowhide. In the sterile -male technique, male insects are raised in an insectary where they are treated

137 PESTICIDES AND PEST CONTROL with radiation to make it impossibly for themto reproduce. These sterile males are then releasedto-mate with *tales who later lay eggs that develop. improperlyor not at all. Releasing sterile males has boNift quite successful in almost eliminating the populationof screwworms in the southeastern United States. Thetechniques involved in the successful control of thescrewworm have rarely been duplicated for °tiller p--sr-sts during thepast 25 years. Researchers are also exploring the use.of.chemicalsterilization, btit at thiktime all of the substancesthatarq effective chemostehlants have harmful effects on higher animals andhuman beings and are 1 hazardous even under carefully supervisedconditions. Unfortunately, it appearsthat the sterile -male approachto insect control (although scientifically interesting and feasible forcertain insects like mosquitoes and fruit flies) is not apt to have much broad impacton reducing chemical insecticide usage in thenear future. At right is a group of blowfly itiaggols rewworms) with oile tiiilawd to show detail Below an adultblowfly.(COLHIO- nflailOMIIIIUraX) 41'1115 1.N) 4

~~138 ONLY ONE SCIENCE~~

144 Hormones and TwO of the most highly publicized methods of "natural" control Pheromones are the use of insect hormones and pheromones. The hormones can interfere with the natural physiological processes ofpest, and the pheromones can affect a pest's behavior. A number of insect hormones have been identified. The hormones are usto kill by.interferingyith, confusing, overstimulating, or inhibiting t fitinsettSitonnal physiologic l functions; they may act on the inse's by chemistry and may affect its gut activity, blood sugar level, or fertility.

One method being used to control blowtlys involves breeding and hatching them m a laboratory where the maks arsteril- 'zed by radiation The resulting sterile male blowflies are then released in an mfested area to mate unsuciess fully, thereby 'educing the blow fly popula- lion in that particular area APHIS USDA

139 PESTICIDES AND PEST CONTROL Pheromones are like hormones in that theyare glandular in origli and act in minute amounts. The big difference is thatwhile hormones are secreted internally and act on the pest's body chemistry, pheromones are secreted externally by one creature and act on another,usu, ally of the same specie's. Pheromones exercise behavorialcontrol over, colonial insects through their odois whichmay cause clustering, indicate alarm, or blaze an olefactory trail for others follow. Probably their best known function, however,is as sex attractants. Pheromones of important insect pests are being isolated and-identified,and some are available commercially. Traps containing sex pheromonesare used to attract insects so their populationscan be monitored. Another use of pheromones in the management ofpests is to suppress mating by confusing the male insects. This procedure is stillexperimental; but it rests on the theory that if many sources of syntheticsex pheromones are introduced into the native environ'ment of Vie insect, there will bea greatly reduced possibility of the maleor female insect finding a member of the opposite sex. Methoprene, produced by ZoeconCorporation, is one of the few hormonal pesticides now available. It is usedas a flood-water tAi) ,t4Cii thewit' 5itUthi mosquito larvicide and also for the control' of hornflies (Siphonairritans). (belOtt, rttyt) can be used to lure Although use of hormonal pesticides and control by ulist

~~BOLL WEEVILS a.~~

~~140 ONLY ONE SCIENCE~~

~~0.1111111.7.: C~~

Crop Resistance A successful method of nonchemical control of destructive insects has been the selection and development of resistant crop varieties. Almost everyone is aware that the development of hybrid seed corn revolutionized farming in the United States so much that yields have nearly tripled in the past 50 yrs. However, few people realize that research to improve s ed corn is still very active. One major goal of this research is to develop ybrid varieties that are resistant to pests and diseases. As an ex ple, the research cenier of Pioneer Hi-Bred Inter- national Inc. grows corn borer moths, Ostrin ia nubilalis, the lar.vae of which can easily destroy corn crops. The company dovys the moths to obtain more than 60 million eggs a year. The newly hatched larvae from theseeggs are used to test different strains of corn for their resistance to this pest. ,. The search for pest-resistant varieties must be a continuing process because living organisms can acquire resistance or immunity to nonchemical as well as to chemical control measures. When pressure is applied that comes close to exterminating an organisii, the threatened. organism may be able to adapt to overcome the threat, particularly when long periods of time are invjaved. The 'process of natural selection of insect-resistant and disease-res,istant strains of crs is going on continuously, and research Such as that described with is needed to keepf he farmer ahead of the p /ests. (

Disadvantages One major problem facing the further development of nonchemical methods of pest control is their specificity. Because these methods usually are effective against only one kind of pest, crops still require chemical or other treatments to handle threats from other species'of pests. The second important issue for nonchemical (and chemical) control methods is the possibility that health or environmental hazards may be introduced which have effects at least as severe as those' methoA they are intended to replace. This is particularly true in the cases of hormonal controls and chemical steriliztttion techniques. A third major drawback in the use of nonchemical measures is the cost of developing and marketing them. Because of the specificity of these measures, a feature that many environmentalists consider desirable, any single product will have a limited market. So,Neven if a prod- . .1. Uccould capture the entire market for th control.of a particular species of pest, its limited demand might njustify_the costs of ) development. 44 kr

141 PESTICIDES AND PEST CONTROL 4 . - ) 147 INTEGfie TED v PEST M A NA GEM ENT The most widespread attempt to integrate nonchemicalcontrols with chemical controls is Integrated Pest Management(IPM). Researchers using the biological, cultural, and chemicalcontrpl phases of Integrated Pest Managementprograms have provided valitable. flowledge needed to attack a pest at itsmost susceptible stage and to determine the circumstances under which the hostis most vulnerable. *- Few would argue about either the desirabilityor the potential of, combining multiple methods of pest control intoan effective Integrated Pest Management system. But the lack ofresearch information on the basic biology of pests andcrops, the lack of established econcimiC4t.hresholds, andthe primitivestate of predictivemodeling and agroecOsystem'analysesare serious deterrents to the growth of IPM. . . Although according to the 1979 Reporton Pe, anagement Strategies of the Office of Technology Assessment IPM appears to be the most promisingcrop protection strategy for the next 15 years, [there are] technological and administrative obstacles that impede the develop- . -nient and implementation of IPM. The technological obstacles lie primarily in theareas of . basic knowledge, delivery systems, and personnel. An inadequate base of knowledge in the biology, bionom- ics and interactions of pests seriously limits the It range of control tactics able for integratijig pest management into a total rop p eduction system. c OTA has concluded that, witiricr.. sed use of IPM, chemical ,.. pesticide use is expected to decrease all crops considered in the 1979 report; supqingly, however, it has also concluded that"the amount of reduction is speculative and may not be nearlyas significant as is often assumed by1some persons. It ismore cerfain that the pesticides that are applied will be used efficiently and effectively." - Because chemical pesticides will continue to be heavily used in the foreseeable future, it is important to discover for eachcrop and each pest how to select the most-specific chemical available, themost suit-.. able formulation and application techniques, the optimaldosage, and the proper time of application. Some such knowledge isp. rt of the experience backgiound of any successful farmer. Butto ge e best results, it is necessary to have information about trendsin pes .opula- tions obtained byereawide surveillance andan understanding o e , .

~~142 ONLY ONE SCIENCE~~

-...... 1 4 b tr. , . or relationships between pest populations, pest damage, crop production, and profits. For example, in some cases, while a pestcKntrol technique may reduce damage by pests, it may also decrease overall production. In other cases, the cost of pest control may exceed the benefits in increased cropproduction, making it more profitable to accept reduced production and to avoid the expense of certain methods. Convincing skeptical growers that Integrated Pest Management is a viable alternative to their prgsent pest-control practices is not an easy job. Many growers have confidence in what they are doing now; many, already faced with serious threats from pests, are hesitant to abandon current control methods for the uncertain services and benefits offered in an IPM program. And in many cases farmers do not see any economic benefits from IPM. Because of the complexity of IPM strategies, a great deal morresearch is needed to develop, implement, and improve them.

PEST CONTROL AND SOCIETY Measuring the social impact of pest control is difficult. The reasons for and results of its use are wide-ranging and complex.-One of the more obvious reasons for controlling pests is to enlarge crop production. If a reduction in the number of pests permits farmers to increase their harvest, the farmers may realize greater profits because their unit costs are reduced, and consumers may pay less for food bec'ause it is in greater supply. Pa§f absences of or failures in pest- control programs have set the stage fo.rtuct tragedies as a plague of locusts devastaf g hundreds of square miles of cropPand, causing star- vationinncontrolled insect populations transmitting disease to millions o people, cawing epidemics of typhus, -palaria, plague, or encephalitis, There a e other, perhaps less-compelling, reasons for using pesticides Farmers may find it necessary to apply pesticides to reduce the aphid population on tomatoes to meet the strict tolerances for the number of parts of insects permitted iriAatsup, even though the average consumer may not be aware of any dontamination\of catsup. Cosmetic considerations sometimes mandate the use of`pesticides. For instance, mite damage may affect the appearance of an orange while having no impact on its taste or nutritional quality. An orange's appearance, however, may affect its marketability=a strong incentive for a farmer to control the mites.

~~143 PESTICIDES AND ?EST CONTROL~~

~~CMS , N .~~

~~Pest-Borne Diseases One aspect of pest control that is often ignoredin the continuing debate over its need is the contribution ofchemical pesticides to world health.~~

*.Prior to 1939 epidemic typhus killedmore people during all the wars in history than all the spears, arrows, bombs, andbullets combined. One type of typhus is transmitted bybody lice, a pest that thrives on the unsanitary conditions associatedwith extended battle. War and this disease seem togo together. 1 Early in W orler01 II, long after PaulMuller discovered that DDT..- was effective in controllinga wide range o(insects, including lice, the allied forces put DDT intomass production. In 1944 DDT was applied directly to thousands of soldiers,refugees, and prisoners to combat body lice, thereby helping to avoid thespread of a disease that killed more than 2 million,people duringWorld War I and its after- math. i Malaria, probably the greatest killer anddebilitator in the world, k,(.. is carried by several species of the Anophelesmosquito. Prior to the widespread use of DDT, control techniquesincluded spreading oilon the surface of water to prevent mosquito larvaefrom getting air necessary for their survival, water management, and usingpesticides such as Paris green and pyrethrumspray. 10. Throligh the use of DDT and otherinsecticides, the elimination of potential breeding places, efforts to improvethe standard of living, and the discovery of better medicaltreatments, the picture gradually changed until the disease wasno longer a problem in the United States. Nevertheless, malaria remaineda problem in other parts of the world. It is estimated that in the late 1940s andearly 1950s there were about

300 million cases annually, with at leastone death occuring from the . disease every 10 seconds. In 1956 the WorldHealth Otganization mounted an international malaria- eradicationeffort using DDT. I 0' Although the program wasvery successful in many countries, it'is si '4 estimated that the disease stillcauses 1 million deaths (mostly in Africa), and there are 120 millioncases per year. On the other hand, the persistence of DDT and the harmful effects causedby its accumulation itt in they food chain contributed to its being bannedfor use in the United States. Other seriaks diseases are also borne bymosquitoes. Yellow fever is,still a maior threat in AfriCa although ithas been eradicated in most urban areas of the New World. Filariasis,a disease caused by a parasite 4. nematode worm and transmitted by mosquitoes,infects an estimated' 250 million people. Mosquitoes . are the vectors of more than 80 viruses I.

~~. 144 ONLY ONE SCIENCE 1~~

~~13' .4 !~~

.Gambusla affims (above right) . *$. is a small fish which is a natural. predator of the larvatand pupal stages of mosquitoes of the genus AriopHeles Gambusia (right) ait being liberated into a pond tt, help ,opitrol the mosquito that attack human beings and cause Clisekses, such as o'nyong -nyong s population fever, dengue fever, Chikungunga fever, and several types of USDA/Center for Disease Control encephalitis. Encephalitis, a disease prepent in the United States, is transmitted to people by birsds.and mosquitoes. In most yeirs there is little risk to people, but when mosquitoes become More abundant.--:after heavy rains for ii4tanceLthe risk increases. Researchers are developing a number of nonchemicalalternatives ts:1 prevent outbreaks of encephal- Ott itis in this country. Many of these techniques could have applications in other parts of the world to control mosquito-borne diseases. There is evidence that the mosquito how is exhibiting incredible wadapiability. Southern Asiaand parts of Central and Soui0 Am'erica'

~~145 PESTICIDES AND PEST CONTROL~~

~~/44 151 a~~

6 are experiencing a resurgence of malaria: Mosquitoesare developing a high resistance to hous'eholdsprays and other methods used to control them. 'In addit'on to insecticidalsprays there are other approaches to _ contxo1 osq0bes. Justas soluepeoplehaVeitereditary resistance _ to certain diseases, Mosquitoes have beenfound that resist laboratory infect them wish encephalitis virus. Suchmosquitoes may be released in ho of eventually replacing those whichcarry the disease. The use of naturaledators (for example, fish thateat larvae) is another control meth .d which is effeQvein certain locations. Biologists first th ught that thevariations inthe number of mosquito larvae from rice field to rice field inCalifornia resulted from toxins from a blue-green algae. Recent research,however, has ghown that when'mosquito larvae brushagainst certain species of flatworms in the genus Mesostoma the larvaebecome paralyzed and die. Studiesare now underway to find out why,certairwrice fieldssupport flatworms and others do not. Theseworms may provide anotherbiological .approach to mosquito control. Parasiticroundworms that can kill larvae may also prove`to be the mosquito's nemesis. These are just a,few methods that havepotential for controlling`- encephalitis- carrying Mosquitoes iriltreUnited States. Many of these methods may have applicationelsewhere for &ntrolling pests which transmit other diseases;enormous health benefits could ensue if the techniques are successful.

Risks gnd Benefits When spray operators become alor die as a result of spraying or spilling pesticides on themselves, theadverse effect§ of pesticidesare immediately evident. In other instancesdetermining that a death has been caused by a pesticide oftenrequires a little more detective work. Accidents caused by careless'storage ofpesticides around small children and even suicides.showup in the statistics on pgsticide poisoning. . Immediate and fatal effectaare fairly easilydocumented. Despite the vast increase iin the availabilityand use of Pesticides, since 1956 direct deaths attributed to poisoningby agricultural chemicals, including pest,' ides, dropped from 152 in thatyear to 34 in 1977. (41'1977 the tota amber of lethalikiisonings from allsubstances in the United States was 4,970.) During thissame period the total population and the number of accidental deaths from all otherpoisonings approximatel) doubled. . . Pesticides may"cause unwanted damageto crops and animals which can octur at the site of applicationor many miles downwind or r ea 146 ONLY ONE SCIENCE downstream from the point of application. Some other adverse effects are harder to iden.tify. In general, it is rare that the cause of an individual case of cancer can be identified reliably. Epidemiological studies of population groups may associate a rise in rate of cancer with a common exposure to a chemical, but such detettable outbreaks are usually in relatively small, immobile, and homogeneous populations of heavily exposed individuals. No such outbreak of cancer has been reliably related to a pesticidealthough the possibility of such a relationship cannot be dismissed. The United States Environmental Protection, Agency has established a Cancer Assessment Group with the responsibility of evaluating evidence that certain substances are or are not carcinogenic. Their deliberations 'are principally based on the results of experiments on laboratory animals and findings in epidemiological studies: . r Problems also exist in assessing the more'gerious long-term or delayed effects of pesticides on wildlife and on the environment in general As already mentioned, certain pesticides persist in the environment and may Accumulate in the fond chain. Few doubt that the potntial exists for serious, long-standing ecological damage, but such damage is difficult to demonstrate in many specific cases and is, therefore, almost impossible to quantify. It is not uncommon for a pesti - cide to kill beneficial parasites and predators of pests as well as the pest itself,"thus resulting in a serious infestation of the same or a different pest at a later time. But this effect is also difficult to measure. Biological control methods may also suppress the population of one pest, thereby, in turn, allowing the population of another pest to devel9p into a serious problem. Effects of pest control measures are hard to cla\ss0 as entirely beneficial or adv. erse. It is easy, for example, to under why the Mormons in Utah erected a monument to the seagulls that saved their crops from a devastating horde of insects. It is somewhatmore surprising to learn that a monument was erected to the boll weevilin Enterprise, Alabama, because that insect forced farmers to diversify their crops, thus benefiting the entire state. Few experts in the field of pesticontrol believe tit is possible for the United States to maintain its current levels of agricultural productivity, let alone to increase production for expariding world needs, without some use of chemical pesticides. Similarly; acceptable freedom from pest-spread diseases, pest annoyance, and the destruction of crops and forest products by pests cannotbe\achieved without some use of chemical pesticides.'On the other hand, the tremendous surge in chemical technology several deCades ago, especially the rapid conquest

~~147 PESTICIDES AND PEST CONTROL 4 ;, / , , t II. 4 I , 'C-,..-...... ,~~

% of serious pest problems during World War II,resulted in an unjusti- fied confidence on the part ofsome people that chemicals could solve all pest problems. Scientistswere less sanguine about having found the ultimate solution to pest control thanthe users were, and the scientists have been proven right. Today the needto reduce damage, health hazards, and annoyance fro peps is still present.. complishing effectivest control will require more research: ehazarca'of pesticides needto be reduced; the ecological interactions among people, pests, 4nd'environment neecio be better understood; the ability of those concerned witspest control to communicate among themselves-needs to be increased; and the technology f6rdeal- ing with the compleXities of the pest-controlequation needs to be improved. Answering questions involving the risks and benefits of pesticides in today's'society is the responsibility of legislators,rtulatory agencies, county and state research and advisoryservices, agricultural chemical:companies, farmers, and fiIly, incrviduals. All liaveea vital stake in the continued successfulma gem of pests.- / The present political reality in the ted States And in other parts of the world is that consumer and otherinterested groups object to the use of a technology unless they have been persuaded that the benefits of fhe technology are great enoughto justify the ris4cs that are involvid. Within the last decade thesegroups have awakened to the - fact that presSure from social forcescan result in the banning of hazardous pesticides. It is likely that thismonitoring process will con'- tinue; it is clear, therefore, that theases of assessing risks, benefits, . costs, and the ways in which they redistributed tovarious segments of society must be better understood.To this end, both experts in the f ield and interested segments of the populationmusk learn how to communicate more effectively with each other. If mutualrespect among the various segments grows, it should be possiblefor adver- saries to understand and agree on the najpre of the risks,benefits, costs, and alternatives,even if they attribute different values to each.

~~4 r~~

~~14.8 ONLY ONE SCIENCE~~

~~r " 4 I ;1 ' V41) . 5.10 U- J 411101141=11,W.N.W.I.,~~

~~mi. we gramma ow , vommormoo.,,,-.. . 11.1.06.11. ro I IL r-17,7,7 " ,..."011.11 IMMO $.1101,1 011/1/111111111m* OM* 4'' 1d ,1 ' r, Synthetic VA VAAVFibers WAY~~

About 30 years ago the term "wash-and-wear" was introduced to the American public. Since then It has become virtually synonymous in the American mind with any fabric made entirely or in part of fibers created in the laboratory. *The easy care that is characteristic of synthetic fibers is already legenclarylover the years their durability has also proved undisputed. s Vt Fpr thousands of gears, fabric had been woven from natural When the Hai ternmial at the neia, juidah international Airport is fiber, such as silk and wool from anlinals and cotton and flax (linen) Completed, it it ill be 0)-Lered by from plants. Not until the twentieth century did scientists and 5 5 Prtilho,T square feet of engineers successfully reach beyond the range of fibers provided by coated fiberglass fabric The translucent tw.m t) nature. Research made it possible to use feedstock chemicals to pro- tent-like twits spatiiiiti 105 duce synthetic fibers, and powerful new technologies provided the acresan area equivalent to that of 80 football fields The term,- means to reform, shape, blend, coat, and treat, the fibers. nars principal purpose is to serve Muslims making then once -in- Today more than one;third of the fiber used in the world and a- life -time holy pilgrimage about three-fourths of that usectby mills in the United States are syn- Mecca Six 'twit/red thousand thetic. Synthetic fibers are everywhereproviding clothing, shelter, pilgrims are expected to use the Moldy Ill 1981This building and household furnishings, irritiroving long-range communications, matccof w as 1,e,awie and helping make more time available for Wisure and for work. Collec- its high strength, resistance to weather, arid long lire tively, in their many apPlicationsTsynthestic fibers make the modern

~~Owens-Coming FthergLo world more habitable. C.~~

THE VERSATILE FIBERS Today, fiber chemists can.produce and tailor synthetic fibers to meet a variety of special needs. Synthetic fibers also canbe produced more rapidly air some cases more economically than natural ones, especially in recthit years. Polyester staple, for, example, is usually cheaper than cbtton, and nylon and polyester filament yarns are

151 SYNTHETIC FIBERS cheaper than wool and silkprimarilybecause of ,the differences in the amounts of labor required to produce natural andynthetic fibers. The textile mantacturingprocesstends to become increasingly labor- intensive as theraw product'goes from fibr to yarn to textile to finished product. Making the fibers,dyeing them, and finishing them represent less than 10 percent of the retailcost of a garment, while spinning, weaving, design, and manufactureaccount for more than 50 percent of the cost. Making cloth from syntheticfibers requires less labor than making cloth from natural fibers.Because of this saving in labor costs, the average family in theUnited States now spendsa much smaller portion of its incomeon clothing than-it did 20 years ago. Synthetic fibers have made dramaticchanges in clothing possible. Today's comfortable, fitted,,and functional stylesare possible only because of synthetic fibers. Synthetics haveeven sparked a number of fashion revolutions. Textured nylon,for example, led to pantyhose, which in turn helped popularize theMiniskirt. Modern styles of lingerie, swim, and other sportswearowe their existence to synthetic -- 77r- fibers.

Synthetic fibers airsomean that not only the wealthy but alsoa vast spectrum of Americans can enjoy the latest trendsin home decoration. The fibers have made possiblea great variety of inexpensive, easily maintained, durable, and colorfulfabrics for upholstery, draperies, and linens.. In fact, bulked,continuous filament yar,ns have creatd a carpeted America. Because synthetics are so easy'to maintain,the need for mending has decreased; and although peoplenow wear more knits than they did a generation ago, darning is almosta forgotten art. In the past, an average family's ironing may have takenmore than 1 day a week, and this was for approximately one-half or less of the clothing andfurnishings that Americans own today. The importanteffect Of these time and labor savings is that people havemore time which can be used in other ways. This time can be used to add to the total workforce, for recrea- tion, or simply for relaxation. By cutting the cost of recreationequipment, synthetic fibers have also helped diversify and multiply leisure-time activities. The private , boating boom is largelya result of research that made possible the development of synthetic fiber andpolymercomposites that can be molded into hulls reinforced with glass fibers.Many of the low-cost, sturdy, lightweight tents, sleeping bags, andother hiking and camping equipment more resistant thanever to extremes of heat and cold, wind, rain, and abrasion could not exist withoutsynthetic fibers. In skis, golf

~~41'152 ONLY ONE SCIENCE~~

~~D t 4~~

~~club shafts, tennis rackets, tent poles, vaulting poles,and skate boards, where strength-to-weight and flexibility-to-weight ratiosare impor-~~

The construction industry is using synthetic fibersto cover some long, clear spans. Sport fans know that the Silverdoine Stadiumin Pontiac, Michigan, has an inflated roof woven of glass fiber.A Teflon® coating proteCts the roof against extremes of weather,yet the roof remains translucent. The flame-resistant qualities 2f glass fiber andits high strength-to-weight ratiowere importantfactsin the choice of these materials. When the cost and timeto erect this enclosed stadium are compared with the cost and time needf4to build similar structures in recent years, the economic advantage of this inflated roofover conventional ones is easily demonstrated. Inflated structures also have permittedsome large-scale construc-. tion projects to proceed virtually all winter long ineven the coldest cli- mates. Such structures also are being used in hydroponic gardening and farming to make plant growth independent of the weather. Experts predict that the ability to erect these structures quickly and \--inexpensively over large area will haveincreasing application in construction, agriculture, and indutries in which pollutants must becon- trolled. Even as they are being further refined, synthetic fiberscontinue to be used for a variety of special environmentaluses. These include hollow-fiber membranes usedto purify water by hyperfiltration and filter bags that can remove particulate matter from smoke.Nonwoven fabrics made of olefin, nylon,or polyester are serving in road construction as a textile layer between road bed and surfacepavement; they

~~The StIverdome Stadium in Pontiac, Michigan has more than 80,000 seats'and covers about 35,000 square meters Owens Commg Fibe,glas~~

~~153 SYNTHETIC FIBERS~~

~~158 4.~~

resist the elements and prevent cr4king of the road surface. A less well-known use of synthetic fibers is in a so-called nylon "whale" capable of sucking up thousands of gallons of oil spilled into the ocean. Not too lortg after synthetic fibers became available, surgeons started replacing portions of damaged or blocked blood vessels with tubes woven of synthetic Fibers. For about 20 years synthetic fibers have formed the soft, woven synthetic fabric thatorings an artificial heart valve, permitting it to be sewnto the wall of the heart. Even the spture materials are frequently synThetic fibers: Synthetic fibers also are used for such nonwoven hospital dispos- ables aS bed linens and gowns for professional and nonprofessional personnel. Glass fibers also are used in probes made of special glass or plastic fibers which can transmit light. These make it possible to see inside many of the body cavities and hollow organs, includirig tie abdomen,'bladder, chest, and the lungs themselves, without resorting to a major surgical procedure to make a diagnosis or to give treatment. In the last few years, medical scientists have been experimenting with the use of carbon fibers in artificial joints becaiise these fibers are lightweight, very strong, and chemically inert. Iiideed, the number and kinds td-medical uses for synthetic fibers have increased markedly in the past decade, but these filers represent only a small portion of the total volume of synthetic fibers used for all purposes because medical devices usually are small.

~~Human blood vessels May be repaired using grafts made of synthetic fibers such as this Veri Soft woven graft (130x) Meador Medicals Inc~~

~~44v 154 ONLY ONE SCIENCE~~

~~15J a. i4ow SYNTHETIC FIBERS ARE~~

MADE . Polymerization is a chemical reaction in which monomerssmall chemical compounds with low molecular weightcombine to forma polytner, a more comple'x,molecule with a higher molecular weight and very different chemical and physical properties. Polymerization is basic tb the manufacture of synthetic fibers: Theprocess may be visualized as the end-to-end joining of molecules, something likejoining the links of a chain. If each link were about' inch long,some polymers might be longer than a football fN.-It is a polymer's high molecular weight and its enormous length-to-width ratio that account for some of its unique properties. Polymers can be molded or shaped, they can be set'by heat, chemical processing, radiation, or by a combination of all three processes,Some are more elastic tan rubger, and othersare so tough they are used to produce helmetsnorbulletproof vests. Production of a synthetic fiber begins with liquefaction of a polymer by dissolving it in a solvitnt or by melting it into a syrupy liquid. It therlis pumped through a spinnerette, a nozzle-like device similar to a showerhead containing tiny hOles. As the filaments emerge from the spinnerette, they are hardened or solidified. This process is called spinning. It is spinning somewhat in the manner in which a spideis said to spin and should not be confused with the textile operation that produces yarn. In the manufacture of synthetic fibers, there are three methods of spinningwet, dry, and melt. In wet spinning, filaments from the spinnerette pass through a chemical bath where hardening takes place. Rayon is produced by this technique. In the dry method, the extruded filaments dry and harden in warm air. Synthetics processed by this method include modacrylic and acetate. Fibers produced from polymers that are melted before extrusion and hardened by cooling include nylon, polyolefin, and polyester. A separate process in each of these techniques stretches the fiber after they are extruded from the spinnerette, reducing the diameter of the fibers in proportion to the amount of stretching. This stretching aligns a substantial fraction of the polymer's chain-like molecules parallel to the length of the fiber, greatly increasing tensile strength and stiffness. Synthetic fibers can be blended with one another or with natural fibers. Two different polymers can be extruded side by side and

~~155 SYNTHETIC FIBERS~~

fit 16o This machine was developed during the early days of nylon produttionit pressed nylon filaments into a loiig ribbon, which was then cut, melted, and extruded to form individual nylon fibers of varying thicknesses Lb 1,114, ra N.111t.IIdiI.Founda- IP loq

twisted toget er, or theycan be mixed and extruded as a single fiber. Sufibers c give the finished product desirable characteristic's,such tistatic, elastic, or flame retardant properties. 0 Synthetic fibers can be made in different shapes and thicknesses; they can be textured to add bulk, stretchedto conform to desirable shapes, or precliac ed in varying lengths for blendingwith other fibers. With special additives, basic fiber-formingsolutions can prodiaCc--'

~~156 ONLY ONE SCIENCE~~

r --_,...., r fibers that are as comfortable as cotton or that look and feel like furl_ wool, or silk. Some synthetk fibers can be manufactured into fabric's that are both sheer and strong at thesame time. The various processes available for the maluailiouture of synthetic fibers and' the wide choice of polymers and molecular weights give these fibers their versatility in meetingmany different needs. 'C , RESEARCH AND -4 DEVELOPMENT / Rayon (Artificial Silk) The oots of fiber technology, like theroots of many early tech-y nologies, re traceable to Europe. The earliest proposal for making artificial ilk is attributed to a versatile English scientist, Robert Hooke. His observations of silk worms and spiders prompted him to boast that .., he could-make a better fiber than thsilkworin. In 1664 he suggested that if a suitable liquid coul be forced through small openings,it would harden into fibers or artificial silk. Hooke, however,.wasa man of varied interests; he did not follow throughon many of his ideas, and he tended to abandon any slow-movingexperiments. In the 1880s, a Swiss chemist, Georges Audemars, alsowas . interested in making artificial silk. He reasoned that the mulberrytree could provide people, as well as silkworms, with thenecessary raw ingredients for making silk.rHe experimented with the cellulose (the chief constituent of the walls of plant cells) from theinner bark of the mulberry tree to form a suitableliquid, but he laCkeda process for extruding the liquid. Britain's Sir Joseph Swan solved Audemars' dilemma. Swan found that he could produce fibers by ford/1g liquid througha small opening into a coagulating bath that hardened the material into fibers But because Swart was interested in making carbon filaments riyi-e 1ec t ric lamps, he never pursued the possibility of making textile fibers.It was Count Hilaire de Chardonnet, a French chemist, whbwent on to de op a process capable of producing artificial silk, later called rayon. tric-t y speaking, rayon is not usually considereda synthetic fiber `because it is made froma naturally occuring polyrrier, cellulose.) Fabrics were first displayed by de Chardonnet in1889 at a Paris exliibi-,, tion; about 2 yearslater he opened the first commercial rayon plantin France. In the meantime three English scientists, C. F. Cros4, E. V. Bevan, ancl'C. Beadle, were working to improve de Chardonnet'sprocess for manufacturing rayon. They developed what is called the viscoseproc- ess, the usual way of converting cotton or wood cellulose into rayon.

~~157 SYNTHETIC FIBERS~~

162 Two Swiss chemists, Carriffle and Henn Dreyfus, patented anoth method, the vetate-preee--s-8 changes-the-rya tare-efreiltilose by treating it with acetic acid and other chemicals. These two brothers made and marketed low flammability lacquers, plastics, d films. Most experts agree that fiber technology came to the'United States in 1910 when the American Viscose Corporation, now Avtex Fibers, Incorporated, began manufacturing rayon in Marcus Hook, Pennsylvania. Acetate fiber was first successfully produced in the United States on Christmas Day in 1924 at a plant established by the Dreyfus brothers in Maryland; that plant was-the predecessor of today's Celanese Corporation. Viscose and acetate rayons were popular with American shoppers throughout the 1920s.

Nylon. In his Principles of Polymer Chemistry, P1 Flory points out that the overriding goal of most chemists atoundlie urn of the century seemed to be the desire to prepare or isolate pure substances; that is, substances made exclusively of a single type of molecule. This objec-, tive continued to dominate synthetic chemistry in the 1920s. A chemical discovery was not accepted unless its composition were confirmed through laboratory analyses and the determinations of its proposed molecular weight were shown to conform to the molecular structure. Although synthetic chemistry produced thousan"of different combinations and permutations of atoms, the horizons of research were restricted. Most organic ajd physical chemists came to believe that every definable substance could be classified in terms of a single discrete molecule that could be represented by a concise formula. Theoretical chemists fOcu,sed attention on "the molecule," postulating laws for ideal molecules which were visdalized as small sphes not much larger than single atoms. According to this concept, the understanding of the molecule would serve to explain allpf the prer- ties of a substance. This perspective, however, overlooked a very important group of substances, the polymers, which cannot be described by conventionaLmolecular formulas and which are than building blocks of natural fibers-. Polymers are much larger than most other chemicals. For example, a simple sugar, such as glticose, has a molecular weight of 180, while cellulose, (a natural polymer) is formed from-hundreds of glucose molecules chemically combined end-to-end. The mysterious physical and chemical nature of polymers took years of investigation to unfold. Until the 1930s several theories about

~~158 ONLY ONE SCIENCE )~~

~~How a Long-Chain Polymer, Nylon-66, is Formed. Hydrogen with a single bond connecteto one other atom~~

~~Oxygen wttwo bonds connected to one or two Other atoms~~

~~Nitrogen with three bonds connected to three other atoms~~

~~Carbon with four bonds connected to three Or four other atoms~~

~~Each type of atom tends to form # a certain number of bonds~~

~~ADIPIC ACID (HEXAMETHYLENEDIAMJNE~~

~~One molecule ofadtpigacidand one mole o f hexamethylenediamme which each atom has the desired number of bonded atoms) are brought together~~

When the two inoleculfrc>brought together and heated, old bonds are broken (ooo)and new bonds (Hare formeto make a water molecule and a polymer fragment One end o former adtpic acid molecule is then able to react further with another end of ahexamethylenedt. nits fragment, thus creatiflg a repeating unit or polymer building block \r?

WATER WATER 11*

~~This process is repeated many times, fragments are connected, and a long-chain moleculeis formed This particular phlymer is nylon-6a~~

~~159 SYNTHETIC FIBERS - o~~

these giant itolecules held sway. 011e was that cellulosewas made up. ..;- of clumps of glucciiiig-gegated insome unknown fashion. Another theory was.that relatively simple moleculescould combine through strongsihemical bonds to form giant moleculesmblecules10 times, or even 1,000 times, larger than the original molecules, The latter-was a . less pOpular theory; most scientists investigating thestructure of polymer,s preferred the concept of finite certainty offered by ring formulas to the vagaries of a formula for,chains of unidentified lengths.

The dive Qtearies about the structure,pf polmers precipi- , tated heated discussi rrs and controversies. Manyconverts to are giant molecule concept were zealous in their efforts to define theprecise structure and properties of polymers. Among the princi1 proponents of the giant molecule theory were Herman Mark, K.' er, and H. Staudinger. Staudiiiger led the parade of convertsto the giant molecule theory,. but many thought he had gone too far when he proposeda relatively gimple relationship between the viscosity ofa solution of a polymer and its molecular weight-ci.size. To disprove the theory, Harold Hibbert sand his students at McGill University in Canadaundertook an extravagant program of synthesizing ever larger molecules by classical methods to disprove Stauditiger's viscosity theory. Their efforts had the opposite result from the one intended. Their research confirmed that polymers are actually chemicals of high molecular-weight andthat the viscosity of diyte polymer solutions is determined bY"the'polymer's molecular weight. (In 1953 Staudinger receiveda Nobel prize. for his Work on giant molecules:) Once it was determinedthat natural fibers have uniqueproperties because they are composed of,giant molecules, chemists beganto synthesize giant molecules from smaller ones and to study their properties.. This type of research becamvne of the iajor activitiesat E. I. du Pont de Nemours & Company, Incorporated, in Wilmington, Delaware. In 1927 the Du Pont ComPany initiateda research program-in organic chemistry. The aim was to gain a better understanding-Ofthe,- chemical processes involved in polymer formation and possiblyto find new paths of applied' research. A year later a brilliant young chemist, Wallace Carothers, w,as peisuaded to leave Harvard University,where he-held the rank of instructor, to direct this workat Du Pont. From the beginning, it mbers of the Du Pontgroup were encourdOcrld'elect theirown projects. Carothers was interested in sdl;s41(ces ofhighmolecular weight and in polymerization by conden- .sadn Through his basicresearch on condensation polymerizatiop, he

~~160 ONLY ONE SCIENCE~~

~~11.1A.J 0 Vie 6~~

In 1927 Wa IlacCarothers-left his instruct° 's position at Iiitrvard UM!, rsity to head a research group at E I du Pont de NernourS 6- Company, Incor- porated, inhere he initiated and guided the reOarch that even- litany led to the discovery and . production of nylon tientheriart Inc

~~161 SYNTH.ETICFIBERS~~

166 obtained polymers, chricalscharacterized by recurring structural units. After investigating the preparationsand properties,of many, such substances, Carothers madea significant advance in their Preparation. He used a device calleda molecular still that had been set up at Du Pont by Julian Hill, a member of thegroup. The equipment made it possible to carry polymerizationcloser to completion. secause the still could elimin'ate water formedduring the condensationreac- tion, Caiothers could obtain linear polymerswith much,higher molecular weights thanever before. Some of these substances had molecular weights of over 10,000; Carothers calledthem "superpolytners." (Today polymers with molecularweights of more than 10 millionare known.) The polymers which Carothersmade in the molecular still were only of theoretical interest, however, because oftheir low melting poirits and easy solubility. One day Hill made a crucial observationwhich dramatically changed thceocu of the research. Hefound he could obtain filaments by using a rod to pull threads fromthe molten polymer prepared in the still. Furthermbre, after these filamehts hadcooled, they could be drawn to several times a-1'0r original length.These cold-drawnfila- ments had quite different physical properties fromboth the polymer , / and the "undrawn" filamtntspulled by minimal tension, from the mol- teRpolymer. The laboratory bqzzed with excitement andactivity as the group explored the properties of 'these cold-drawnfibers. X-ray diffraction ". patterns showed"that the filaments in the undrawnstate were partly crystalline and that the crystals hada random orientation. The cold- drawn filaments, however, indicateda considerable degree of orientation parallel to the fiber axis. This patternwas similar to that shown in. natural silk fibers or in rayon filaments whichare formed under tension. The new filaments hact high tensile strengthand elasticity. They- were pliable and tough enough to be tied into hard knots. Drawingthe fibert also caused them to developa high degree of luster and transpar- ency. Moreover, unlike ordinary textile fibers, their wet tensile strength was no less than their dry tensilestren Until this time Carothers' research had beenent rely fundamental, conceived to explore certain aspects of thepolymerization process. However, the striking qualities of the cold-drawnfibers obtained from the polymers aroused hope that it might be possibleto make a fiber with commercial utility A new research phase beg,an at the Du Pontlaboratorythis' time to synthesize a poixter that might form the basis fora new marketable fiber' in his earlier research, Carothers h of been successful in

~~162 ONLY ONE SCIE4CE J~~

A synthetic fiber was first drawn tmmtatest tube in 1930 by Juhan Hill, a member of the research team at D14 (but ( Hill reenat ted that hiatork Moment0pho- tographers in1940 )Ftve years of intense xesearch waa required to produce a marketable p:od- uctrrylon.?0 Lleutifd \fill, th),

~~163 SYNTHETIC FIBERS~~

~~163 I~~

concentrate on a particular polyamide which, when treated in the molecular still, had a melting point of 150° Celsius. Aftercold-drawing, these fibers were equal to silk in strength and pliability. After this discovery, Carotheys and his colleaguesprepared polyamides from a variety of substances and,on February 28, 1935, the first superpolymer from hexamethylenediamineiand adipcidwas synthesized, The resulting polymer, poly(hexamethyleneadipamide), Was designated "66" and later callednylon66. The first digit indicated

the number of carbon atoms in the diamine, and theseco,nd digit, the, number of carbon atoms in the acid. Thenew polymer 66 produced fibers that were insoluble in common solvents and meltedat about 260° Celsius, providing a margin of safety above usual ironingtempera- tures. Du Pont selected this particular polymer for development and marketing because it had the best balance of desirableproperties and had a low manufacturing cost:

~~The third phase of research in the laboratoryat Du Pont was focused on developinge process for themass production of nylon 66. To implement this Process, itwas also necessary to develop the chemi-~~

m. cal and engineering know-how for the erection andoperation of a large scale plant These tasks wereenormous and required the time, talent, and ingenuity of some of the best chemists andengineers of the time. Processes for the manufacture of adipic acid and hexamethylene- diamine had to be workedout. The machinery and proceduresfor cold-drawing, sizing, twisting, and packagingyarn had to be planned. In the process of working out all,of these elements ofthe operation, many difficulties were encountered. No previous experience with these types of processes existed in the DuPont Company,or anywhere else for that matter. Nylon polymerwas a completely new material with properties different from those ofany other previOussynthetic product Spinning nylon from molten polymerwas completely different from spinning either acetateor viscose rayon.

One example of a complex problem which neededto be solved . Was the development of thepumps to handle the viscous molten poly- The spinneret was not a new mer "Special pumps had to'be constructed whichperformed under ma IiinrIbenitwasuseduitbt severe temperature conditions, had only small clearances, and used production of nylon. It did, however, have to be redesigned only the polymer itself as the lubricant. A special abrasion-resistant au rtb.dattNome ofthe steel which did not 'warp or soften under these conditionswas neces- tong Lie rec}ttrentett esof nyloir p7odp, hop sary. In fact, the entire spinning assembly involved radically new engineering accomplishments to produce fibers of the requireduni- formity and other,desirable qualities.

~~164 ONLY ON 'SCIENCE~~

e American consumers were, at first, unaware that they-were using nylon when, in 1938, toothbrush manufacturers began substituting it for Chinese hog bristles. But Americans awoke to the promiseof synthetic fibers at the Golden Gate International Exposition in San Francisco in 1939 when nylon stockings were shown publicly. Buying was frantic when the stockings first went on sale nation-wide about 1 year later. During the first year, retailers sold more than 60 million pairs. The major advances which allowed'the production of nylon did not benefit the American retail market for very long; when the United States entered World War II all American nylon was allocated for the production of war materials. Japanese silk was replaced by American nylon for manufacturing parachutes and government currency. Americapeclothes made from nylon soon disappeared from the market because ito synthetic fiber was needed for tires, tents, ropes, ponchos, combat clothes, cargo rigging, and other military supplies. Although cotton still made up 75 percent of the national fiber market by fhe end of World War II, there were indications of change. The American public wag buying all of the nylon stockings it could, ancVnylon was being introduced into carpeting and'automobile upholstery.

Nylon stockings, first demonstrated early in 1039, were made using experirtiehtal production techniques The full-scale p,o- ductron of nylon u not begun . by Du Pont until eceitibe,, 193 9 .4 Drittite,hyr MIN Harry r,14,41,7 lit,Ili,

~~Ts. 165 SYNTHETIC FIBERS~~

~~170 a, 4 ?Du~~

~~r- 11' Each of these 1 1000 bobbins a., represent a different item in the 44.4 nylon product field Eleuthertan tftlis-Hagley er,..", n~~

~~14,~~

Other Synthetic Fibers Not every, technology durrently in Americanusage came from- theefforts at Du Pont. Much of the fibertechnology had its origins in other research bases established priorto World War II. In 1893 Edward D. Libbey displayeda fabriC containing glass filaments. Scien- tists from Owens Illinois Class Company and Corning GlassWorks combined efforts to develop a method of spinning relativelyfine glass filaments, which eventually resulted ina product that was excellerit for insulation and for reinforcing plastics. In 1938 thesetwo companies

166 ONLY ONE SCIENCE formed Owens-Corning Fiberglas. orporation. In the early 1960s Owens-Corning produced a very fine glasafiber that was pliable enough to be used in fabrics. The American Viscose Corporation, thisnation's first manufacturer of rayon, was also the first company to spin vinyon into textile filaments in 1939. This fiber, produced today by Avtex Fibers, Incor- porated, and the Union Carbide Corporation, is remarkably inert to high concentrations of alkalis and mineral acids. However, it is hard to color and has a relatively low melting point. Since 1946 when they were first introduced by Doebeckmun Company, metallic fibers (those composed of metal, plastic-coated metal, metal-coated plastic, or having a metallic core), have been used as decorative devices for clothes and home furnishings. Srletallic fibers such as LurexI are found in items such as braids, military urtiform decorations, draperies, laces, ribbons, table linens, upholstery, and in even carpets (to reduce the build-up of static charges). suitable adhesives male it possible to add color to the fibers; and, when properly treafed, the fibers are not affected by salt, chlorinated water, or electrolytic or climatic conditions. In 1948 Union Carbide introduced modacrylic fibers These fibers have been used to simulate fur and are also found in awnings, blankets, carpets, curtain, scatter rugs, filters, paint rollers, and stuffed toys Modacrylics remain popular, primarily because they clO a good job of emulating the soft, yet resilient, easy-to-dye properties of yvool. They are also flame-resistant, quick-drying, and keep their basic shapes under most conditions. In the early 1950s Du Pont and a few other companies began producing acrylic fibers AcryliC production represented the successful development of a synthetic fiber which simulates certain charactecis- tics of wool. Acrylics can be used to make a variety of products includ- 4 ing spo tswear, infant wear, sweaters, work clothes, blankets, carpet and yas for hand knitting. , Ac Tics and other synthetic fibers that are flexible enough to be joinedith other materials made blending with cotton or wool possible. The first result of this mixing was the "wash-and-wear" apparel that appeared in 1952. The original blend was alratio of 60 percent acrylic to 40 percent cotIon. The aim of the blending process was to couple the best characteristics of cotton with those of the synthetic fibersparticularly since it seemed to many experts as though some of the basic advantages of cotton, such as moisture absorption, could not be duplicated.

167 ' SYNTHETIC FIBERS Polyester fibers, whichnow represent the single most important class of synthetic fibers, also originatedfrom Carothers' pioneering research at Du Pont. Carothersdiscovered their fiber-formingproper- ties but, because he was workingwith a particular type of polyester which he found to be too tomelting and too solvent-sensitive, he redirected his research tow_ad thepolyamides (nylon). Subsequently, in1941,J R. Whinfield end J. T. Dickson,working for the Calico Printers Association in Great Britain,found that polyesters basedon terephthalic acid, an "aromatic" acid,produce very desirable fibers. I Their basic patent was acquiredby the British company,.Imperial Chemical Industries:for all,countriesbut the United States. Ina strange twist of fate, Du Pont acquiredthe United States' right to the commercial fiber whichwas based on fundamental research done in its ownlaboratories. Polyester fibers became commerciallyavailable in the United States in 1953. The rapid growth intheir production was aided in the 1960s by false twist texturing,a prdFess in which fibers are crimped (or convoluted) to decrease contact with theskin and increase air trapping thereby increasing comfort. Falsetwist texturing made the popular shape-retaining polyester knits (typically65 percent polyester and 35 percent cotton) possible and, in addition, remarkabryincreased the use of synthetic fibers in tuftedcarpets. Another significantuse for polyester is as fiberfill for pillows and , "cold- weather clothing. Polyesterfibers have largely replaced nylon in I tire cordyncause tires containingpolyester not only are strong but also are not subject to theannoying"flat-spotting" that is characteristic of nylon-reinforced tires Intensiveresearch has produced a gamut of polyester fibers witha broad range of properties, includinga type of polyester fiber that can be slowlydegraded by the human body and that may, therefore, be used forsutures that need, not be removed In1959Du Pont marketed spandex,an example of a successful "engineered" fiber. It was designedto meet themand for a stretch- ableyarn to be used in items suchas swimsuits. A stretchable yarn was needed that would be superiorto those thekn available, clearor white in color, easily dyeable, resistantto fading, and stable whin exposedto the chemiclls added to swimming pools.Later, scientists at U S Rubber Company (now Uniroyal,Inc.) developed a spandex in which polyester and urethanewere joined. . -.....5 Du Pont also introduceda unique polymer (an aramid) called kevlarz which hasa combination of high strength and toughness. never before produced or found in nature. It is usedin bulletproof

~~168 orcrOvoNE SCIENCE A~~

vests,*military helmets, and similar protective clothing. The extraordi- nary strength of the aramid fibers could not have been achieved without a detailed study of the behavior of the aromatic polyamides when they are dissolved to form solutions. The basic polymers for these fibers are different from the nylon polymers because their chemical constitution makes the giant molecules extremely rigid. A great deal of research was required to find solventsat are suitable for spinning fibers; it is the behavior of tie arknid pol ers in the,solvents that is remarkable. As more and more of the pol er is dissolved,,the molecules; because of their rodlike structure, cannot be accommodated in the liquid medium in the typical, random fashion. Instead, when a critical concentration is reached, the molecules rearrange themselves into a crystallike order (a property normally observed only in solids) so that they are more or less parallel to each other. It is this liquid crystalline solutiorrthat is spun to forin aramid fibers. T3ie molecules emerge from the spinnerette welhaligned with each othei and are responsible for the enormous tensile strengths in the resulting fibers. The liquid crystalline `behavior of rigid polymer molecules is still all incompletely understood phenomenon, and there are a number of studies underway to'correlate the molecular structure of rigid polymers with their properties. Anoth Du Pont product is a remarkable-pplymeric material, Nomext, vach has low flammability and can be used at higher tem- peratures than most polymers, It is used for ironing board covers, aircraft upholstery, the outfits of professional race-car drivers, and flight suits for aircraft pilots. The properties of Ke-lar make it an ideal material or use in ii,zliter bullet-proof A z5- magnum bullet was repel:ed b., the ,trst icier of this

~~Keular z aet )~~

To demonstrate the amazing strength of their product, Du Pont had a 280 pound wrestler sit on a trapea suspended above the main street of Leicester, Eng- land by two strands o f Kev tar' Each strand was less than 1/8th of an inch in diameter ,

~~E I du Pont de Nremours it+ Com- is pany in,nprorated~~

~~169 SYNTHETIC FIBERS~~

174IC OP. N itHercules Powder Co. (now Hercules,Inc.) began producing textile-grade polyolefins commerciallyin 1961, although they had been used earlier for specializedpurposes. Polyolefins, derived from ' propylene and ethylenegases, are characterized by a resistance to ,moisture. and by chemical inertness.Propylene-derived fiber is used for general textile applications. 4 Ophcal Fibers The 1960s werea periodiiTtohichprodualibniacilities of thesyn- thetic fiber industry expandedand efforts that began a full decade earlier tp "engineer" fibers continuedwith renewed dedication. Opti- ' cal fiber technology became importantin the 1970s. A number of techniques developed for the manufacture ofsynthetic fibers were applicable and helped speed the productionof optical fibers. Many key studies that have improved opticalfibers related to analyzing the transmission properties of glass, devisingbetter glass formulations, and perfecting commercialscale operations. Mechanical, electrical, . and chemical engineersas well as physicists, chemists, mathematicians, and other scientists Kialve all madecontributions to this technology.

. Fiber optics is that field of physics whichdeals with the transfer of light from one place to anotherthrough long, thin, flexible fibers of glass-or plastic; theseare called optical fibers. Optical fibers have the ability to transfer light aroundcorners because the sides of the fiber i reflect light and keep it insideas the fiber bends and turns.rn In 1870 a British physicist, JohnTyndall, discovered that light could be guided througha clear substance. At a meeting of the Royal Society for Improving Natural Knowledge,he set up a container full of water with a hole,punched in the sideof the vessel to allow' astream of water to flow out. Tyridall shone i a light down into the container and demonstrated that, as the lightcame but of the hole, it was guided along the stream ofiivater and litup the spot on which the water fell. However, it was not until the 1950s thatpeople began to use thin strands of glass to transmit both lightand images over short distances. The-gIass fibers available in1960 could transmit signals clearly for only about 100 feet. Forpractical applications, fiberswere needed which could carry the light signalmore than a mile before needingan amplifier. Obviously, the lowerthe signal loss, the greater the distance could be between the amplifiers, resultingin increased economy of the system. . . The ability of the fibers to transmitlight was enhanced by adding il a coating of special glass called cladding whichkept most of the light i? 170 ONLY ONE SCIENCE

~~1 75~~

' "--, waves inside by reflecting them toward the center of the fiber as they passed through it. Researchers also found that the fidelity of the signal could be improved if the fiber were constructed so that its index of refraction (capability to bend light) changed from its center to its outer edge. The compsition of the fiber itself also presented problems. Materials research showed that impurities in the glass produced transmission losses. Metallic ions such as iron were particularly troublesome in this regard, and during the1970shighly refined processes were developed to eliminate such impurities. Although improvements continue to be made, the losses in signaj4i,e.liprIAW' enough and the quality of transmission good enough hat optical fibers can be used for long distance transmission. Bell Laboratories, Western. .1' Electric, Corning Glass Works, and International Telephone and-Te162' graph have all been involved in making these technological advanees. Optical fibers have many uses. Tliey have replaced small electric, lights in some medical instruments. Physicians use a bundle of fibers fitted with a lens and an eyepiece to view internal areas of the body. Many dentists have drills which are fitted with optical fibers to concentrate light on the tooth being drilled.In industry, optical scanners are built into data processing equipment to detect the rectangular holes punched in computer cards. Optical fibers have been adopted by novelty makers for decorations Bundles of fibers are mounted in a plastic or nietal base withX white or colored rights shining in the base. The light travels through the glass strands and appears as bright,points at the tips of the fibers, thus forming designs.

~~1_111U1_ JUlJUL 41,~~

ELECTRICAL OPTICAL FIBER ELECTRICAL Irt anTpfr.7' t elf',ft 71,7,.. SIGNAL SIGNAL 'lat.. are .t7 !fit,tl,t.77t INPUT OUTPUT m tr,7,....717ted t' It the tll'etthe hitht 7;7..7 LIGHT LIGHT t 011( k wto an elec SOURCE DETECTOR tr77 k.ep mostt the tv,tht insult fire t the, tit t Ii ht wane CLADDING mwt,i its tl tre1a. the';/ Pass CROSS SECTION OF AN OPTICAL FIBER

~~171 SYNTHETICFIBERS~~

170 Currently, however, thegreatest potential applications of fiber optics are in com ications. It is believed that optical fiberscan provide solutions toman of the problems plaguing conventional electronic telephone transscion. Glass fibers' advantageover copper wires lies in the differinnatures of electronic and optical communication systems. Electricc ent used for conventional transmission also creates magnetic fields which distort signals.Other disadvantages of copper include electromagnetic interference,crosstalk, and the need to be amplified at 1-mile intervalsby "repeaters." e Because of technologicalimprovements, optical fibers havevery recently become competitivewith conventionalcopper cable for some purposes. American Telephone and TelegraphCorporation, General Telephone and Electronics, andother telephone companiesnow find that it is economically feasibleto install glass fibers in their systems, particularly to connect aitchingcenters in urban areas, A major advantage to these compAiesis that the optical fiber saves space. A thin fiber cablecan carry many more messages thana thick copper cablecan. Therefore, telephone companiescan add message-carrying capacity to existingunderground ducts without having to dig up streets toput in new conduits. In fact,an optical cable with a diameter of 13 millimeterscontains 144 fibers thatcan carry 43,680 telephone calls simultaneously.To carry almost thesame

~~4r, fi~~

~~Examples of optical at'. cables` for various uses Swot Optical tables, Inc~~

172 ONLY ONE SCIENCE number of calls electronically requires 4 conventio ;al cables comprised of a total of 7,200 copper wires. Optical fibers have other uses. Elect utilities are starting to employ them for communication and ontrol circuits within and between generating plants. Because transmissions using optical fibers are relatively immune from "static" associated with high power electrical transmissions, optical fibers are particularly useful near pow lines. Before long, optical fiber deviceg are expected to appear in UNited States-made cars where they will carry control signals to auxiliary equipment. Cable Television companies are using glass fibers on some main circuits. A two-way television system which presently is connected by coaxial cable is being tested in Ohio. The system is called QUBE; it allows people to respond to questions asked over television and to express their opinions instantly from their homes. The high capacity of fiber optics could make it feasible for such a system to be in every home (A few potential uses of this system are briefly described in the chapter, "Survey Research and Opinion Polls.") More and more, optical fibers will make their way into wiring sys- A tems of aircraft because of their light weight; the armed forces are enthusiastic about this exotic material for many uses. For example, the United States Navy is investigating the use of fiber optics for com- municating with submerged submarines. Optical fiber?, linked with a radio buoy, would allow messages to be sent without the submarines having to surface. The armed services are also interested in optical fibers for telephone lines because a directlice must be jnade to tap light signals, producing a signal leak that can be measurecrand detected. Copper wire, on the other hand, can be tapped by with- drawing signals from the electromagnetic field that surrounds the wires; optical fibers produce no external electromagnetic fields. Although the total sales of optical fiber cable and associated equiprrlent currently are relatively miniscule (an estimated $40 million a year), the principal manufacturers expect business to growl() several hundred million dollars a year by 1985

THE PROMISE OF SYNTHff TIC FIBERS Thlreis much that remains.to be done if synthetic fibers are to fulfill society's future needs. Although petroleum prices have sky- rocketed, petrochemicals are still the most economical starting material for making synthetic fibers. Approximately 1 percent of the United

~~173 SYNTHETIC FIBERS~~

~~I. 7J 2~~

Slates' oil and natural gas prodUctio,L-i.s used to make synthetic fibers. (All synthetic materials consume around3 percet of the prodUction.) About onerhalf eif that amount is usedas rawInaterial for the produc- e, tion of the fibers; the other half is used forenergy for the conversion process . In a world that is dependentupon finite energy resources, how will the fiber industry prepare for the future? Petroleumwill neve again be ieg.Aded as the inexpensiveenergy source that it one . Rising worldwide oil prices, have influenced the direction of fiber-research away from that1 modifying established fibers. Now researchers aggressi pub e ways to replace petroleumas a raw 'aterial in the product'of synthetics with substances suchas coal a d biomass nic resqutces, such as-seaweed, terrestrial 'Pia ts, ga /_ge, and huan nd animal wastes). Although the impact vof tco' petition for of and gYs on the synthetics industries is not - clear, a *experts prediany inexpensive alternatives to these 4ub- stanc s the source feedstock for synthetic fibersany time soon Because of the veryrapici rise in the prices ofpetrochemical feedstock in recent years, som`e people have predicteda for rever- sion from synthetie fiberstlo natural onegitiowever,there are few sighs of thishappeningt One reason is thatmany natwal materials hdve also been affected by energy price increases.Another is that synthetic fibe'rs have such widespreadacceptance and use in industrial nations Natural fibers,as currently produced, often Tecluire a large input of,engy,much of it derived frini the same.fossilfuels needed to provide f6edstock for the synthetics. Cotton isone ofhe most, energy- intensive crops growLai.gequaritities of petrochernically based pesticides and fertilizer.needed to grow cotton, and therefore the cost of this fiber has beertdirectlypaffectedby pKice hikes for diesel fuel. The spinning of cottonyarn and weaving of cotton fabrics

:4 are highly mechanizkd in developed cpuntries. Ina large part, the increased productivity in cotton millsover the last three decades has been made possible by te substitution ofcheap energy for laborers who used to do much of the work 117 hand, The production of textile fibers providesan interesting illustration of fhe comparative energy requirements ofsynthetic and natural materials cotton'and polyester frequently,may be substituted for one another, and their relative use is often determined byprice. One study completed in 197kco9'cluded that 25 percent mbreenergy was reipited to make a polyesttiotton blend shirt than to make°Ike entirely of cotton liowikver,lcotton/polyester, fabricsare mbre durable and require less maintenance than all- cotton fabrics The life cycle energy

~~p 174, ONLY ONE SCIENCE~~

~~JP I~~

~~41.1.101111.11~~

~~410" .4,4111110:67 orle. ,44 .....44,4,41 ifirAitilif~~

~~V F o .1,401-4~~

~~4 f"--~~

~~This iompartsori of 3M'Sfhin- siiiiii.'t It farid polyesttr 4t1,4~~

~~175 SYNTHETIC FIEARS~~

160 .41 tive devices to filter out,polluting palticles.These filters will permit other reserves, suchas coal, to be used' more readily, with less fear of environmental contamination. Many of tlitresearch and developmentactivities concerning fiber-reinforced composites alsoare directed toward finding ways to save energy. Generally speaking, thesesubstances are masses of fibers dispersed,in a polymeror metal matrix, and they are becominga new class of.engineering materials.These produ,ct5,attempt to reproduce S. the intricate structural Networksfound in nature which provideso many astounding functional characteristics. .4 The automobile industry istrying to produce a new generation of vehicles that have significantlyimproved fuel economy. One keyto achievingothis goal is to developstrong, economical, and light-weight materials to use in automobiles. Itmay be possible to reduce the weight of a vehicle more than3P percent by using parts made of fiber- reinforced polymers. . The reduced weight advantagesof compdsites are Particularly important in high-performance aircraftand in military andspace vehicles CommerciaLaiicraftnow-are-being refitted-with theselighter materials to enhance fuel efficiency.For example, the next generation of Boeing aircraft (the 757 andthe 767) will make significantuse of graphite fiber and graphite/Kevlarfiber reinforced composites. In addition to research prompted by energy shortages and the . search for synthetics with the desirabteproperties of wool, cotton, and silk, there are other drivihg forcesfor-more study in the field ofsynthetic fibers. Static electricitytinclothing and carpeting made from synthetics is stilla problem. Methods ofsynthetic fiber producecan be refined to giveeven greater savings to the.consumer. Dyeingand finishing processesas well as soil resistant and repellancy qualitiescan to improved. The expandeduse of glass fibers in telecommunications will require-investigation andtechnological innovations. ;Therrrare.new methods whichmake textiles resist burning.More and stronger regulations Will undoilbtedlyincrease the consumption of flame retardant materials andwill encourage-studieson how to reduce

flammability of 'synthetic fibers. , The synthetic fibers industry hasbeenfirmly foundedon basic research discoveries. Thescience of giant molecules,or polymers, is a relatively young field and is expandingin the United States, Japan, West Germeny, Great Britain,France, and the Soviet Union.Although rnany,of the key discoverieswere made by industrial scientists, the base of research idpolymer sciencein the United States is shifting

176 ONLY ONE SCIENCE . toward univ rs sSeveral large coordinated programs of polymer research have been established at the University of Massachusetts, Case Western Reserve University, and the University of Akron.ter universities such as the University of Wisconsin, the University o Michigan, Massachusetts Instinite of Technology, the University of Utah, and the Polytechnic Institute of New York are major centers of polymer science and engineering. f Currently, the synthetic fiber industry enjoysa promising future, witnessed by the strength of the American companies mentioned in, this chapter All are publicly held and traded corporations, and the.rate and volume of investments in these firms provide some indication of the general health and public co1Tidence enjoyed by the industry. In 1976 investments in synthetic fiber plants in this country totaled $8 2 billion. Synthetic fiber sales jumped from $5.4 billion in 1977 to $7.4 billion in 1979. In 1975, some 190,000 employees, at an average salary of better than $16,000 a year, were making and marketing synthetic fibers The United States still leads the world in both synthetic fiber production and consumption, but Japan and some Western European countries have patterns of synthetic fiber research, development, and increases in production that resemble those followed by the United States in years past. Synthetic fibers now represent about one-third of the world fiber market and their consumption is increasing annually at the rate of about 9 percent worldwide. There is little doubt that the world market already reflects the many advantage\of synthetic fiber's. And given the challenges of the future, synthetie--aer`technology is bound to increase

~~177 SYNTHETIC FIBERS S~~

~~4 No~~

fer This is the ,I.st radtpraph o' an mine adult body taken at a safe.' ezro tire' The sitbie,,uv1 7. exposed to k ray, for 30 minutes Her MY,, ,ie,ktake, b.acelet and hRth-button boots with heels are t !early t,t- bleThe radt,-;gratth was take" nr18 414 J klortori, the 5.0,1 ofVi awn Morton who In lti,to ttr.t iced ether tdr 4tirV,a/ avesthecia Nattc nal Lerar liedRole 0

~~178 ONLY ONE SCIENCE '1~~

~~.1 f Xrays'for Medical Diagnois-~~

Many people think of radiation as a relatively new phenome- nona product of the nuclear agebut radiation has been around as long as the universe. In fact, sight is possible only because of light rays, which are actually "visible" radiatio; heat from- the sun isanother type of radiation. Everyone is fam lar with a rainbowthe beautiful array of color resulting when sunl t is split as it passes through droplets ocwa ter. Just as the colors of > e rainbow are constituent colors of visible light, so a number o o hea ons are constituents of the . nwh 'broader range or spect m o.- called electromagnetic radia- fiats shown on page flit:). Radi ayes, which are at one end of the spectrum, have very different' characteristics. from X rays, which are at the other. But electromagnetiedwaves have common properties. They are all rapidly fluctuating electric and magnetic fields that travel throtigh space with the same velocityabout 186,000 miles per second, the speed of light. Electromagnetic. wave's differ in hqw they vibrate as they move away from their sources, and they also differ in energy. High-energy waves vibrate more frequently than low-energy waves. The wavelength, ordistancebetween corresponding parts, of high-energy waves is shorter than.that of low-energy waves. In other words, the higher the eikergy, the-shorter the wavelength. Electromagnetic waves carry momentum and can transfer momentum the same way particles do. In many circumstances, radiation can be thought of as a stream of particles, called photons, traveling at_the speed of light and exhibiting wave-like phenomena. For comparison, visible light photons have energies of from about'2 to 3 electron volts, but X-ray photons ha-ye energies ranging from 125 to 1,250,000 electron volts. The photons at this high-energy end of the electromagnetic spectrum are called >c rays

~~179 X RAYS FOR MEDIGAL DIAGNOSIS A 9 4~~

~~.1 ,~~

riwilavam f.he« t,(' , "hq and gamma rays. The only different6 ,lef.1,,..tritrI7 11!14,4,1te, the between X rays..andgamma rys- drHer,11,. oh- he is in the way in wgch theyare formed. In most instances, gaintharays 11..,,, -,:,,;,:,,,,,,i ,,,,,,,,,ht. are emitted by spontaneous irrtranuclear ,,,,1,f!Ititl,;k 'ay. decay of radioactiveisotopes; and X rays are producedby transition in the inner electronenergy lev

~~r . Iso ONLY ONE SCIENCE~~

~~tit~~

Cle,...... __ i els 'of_an atom and are usually induced by electron bombardment of a 'metallic target in an X-raylube. Once a photon is generated in either mariner there is nothing to distinguish its origin. The practice of medicine has/been greatly in uerped by the application of X rays"to the diagnsosis(of illnesses and njuries and to the treatment of certain diseases. A roZmately two:thirds of the popu- latipn of the United States is exp sed to X rays for medical or dental purposes each year. This chapter wi ncentrate on the uses of X rays in medical diagnosis. It will describe the dis'covery of X rays; the developmeyt of some of their diagnostic uses; several experiences of ' scientists and clinicians with X rays, and a iew r9ile.stones in the evolution of,the technology of diagnostic X-ray equipment. Some hitorcical, material in this chapter is based ondescriptions in The Rays, a book' pre'pared by Ruth and Edward Brecher, at the invitation of the Ameri- can College of Radiology Foundation. The use of X rays in medical diagnosis isboad subject. It is, therefore, beyond the scope of this chapter to include discussions about radioisotopes which emit gamma rays, the therapeutic uses of. _ radiation, or other important methods of diagnostitimaging,such as ultrasgnography,;thermography, nuclear magnetic resonance (NyR) dynamic spatial reconstruction (DSR), or positvn emission transaxial tomography (PETT). Although the long-term and short-term hazardg of excessive radiation are also very important subjects;they will not be a primaryvfocUs either. Hazardous effects are well-recognized, and prow cedural and technological Progress has been made to minimize the dang6rs to millions of patients, to those who useXrays in theirwork, and to the public. For example,sa patient getting a routine diagnostic. chest X-ray photograph in the early 1940s would have received a radia- tioddose more than 10 times that to be expected today. The reductions o'closages used in fluoroscOpy Have been even more impressive, in the same time period, new procedures have made it possible to reduce radiation exposure during fluoroscopy to as much as 1/100 of the(. exposure that was common in the early 19405 DoSages to the ovaries or testes, when other parts of the body are,being;crayed, have been reduced more-than 400 times. All of these reductions in dosage were achieved during the time in which'technological imprOvements accom- panied improvements in the quality of the X-ray films. Scientists, clinicians, and the genfaFpublic all recognize that it is important to, keep human exposurito radiation at the lowest possible level, and a great deal of thought,: research, and legislation continue tobe directed toward that goal. o

~~$ ' 1 .-1.el 4 X RAYS pRNEDICAL DIAGNOSIS ) .i, . :.2...k..."-,,,r,/ d 1 ..,-- ..',4. ..,'~~

Y -,...... 1.1.* "`1.... HISTORY oodspeed Jennings, and In 1890, nearly 6 years before Xrays were discovered, two men; rookes J . each 30 years old, met ina physics laboratoryin Philadelphia. Arthur Willis Goodspeed was a professor of physics;his visitor was a photog- rapher, William Jennings. Jennings had broughthis photography equipment with him to the laboratory. Fdli-some time he had wanted to experiment 6y making photographs using light froman electric 'spark, t and Goodspeed had agreed to help him. The two Men set up an induction coil (toconvert low voltage into high voltage) which was connected toa spark gap. Jennings placed his photographic plates so that light from the sparksgenerated by Goodspeed's equipment would make it possibleto photograph coins and other items placed on the plates. Goodspeed'sinduction coil was the one he ordinarily used to demonstrate howa "Crookes tube" func- tiond A Crookes tube, devised in 1875 by Sir WilliamCrookes, wad 'taw essentially a glas.s tube from which most of the air hidbeen evacuated.The negative terminal of a high voltagesource was con- Sir IA It,am Crouke4 nected to an electrode (the cathode), and itspitive terminal was 11532-1019) attach.e.d.1.0..a..n.ather_electrode (the anode). Whe 4 me, u a,r College aitadieleqr current was applied, a stream of electrons, usually referred toas "catho e fays," flowed from the cathode to the anode and made the glass of the tubefluoresce..

~~The Crookes tube was devlsed .n 197: by Sir William Crooke 44,,,,4.1, 'lege Rad, IC~~

~~18Z ONLY ONE SCIENCE~~

.. , iThat evening, after the photography session was over, Goodspeed o demonstrated several of his Crookes tubes to Jennings. Neither man noticed that a stack of unexposed photographic plates, left over frotheir earlier experiments and still wrapped in light-proof paper, wa ing nearby. .Two of the coins which they had been photographing lay on top of the stack. When Jennings left the laboratory, he took both the exposed and the unexposed plates home with him'''. Several days later, Jennings told Goodspeed t/t something mysterious had happened. He had noticed "fogging"f some of the/6ex- plates. When he had developed.one of these platts, it showed the outline oc two unexplainable small round objects. Neither man could account for what he saw, and the matter was forgotten for nearly 6 years..Jennihgs, however, was sufficiently curious about the disc- shaped images to file the plate safely away. In fad, that X-ray plate from February 22,1890 is the oldest known North American X-ray

~~plaiefor which a record survives.~~

~~le-~~

Arthui VN tills Goodspeed c X- ray photograph made ort Febru- ary 22 18Q0 ',..itional Lthrar 0,. Medi, me 4 Crookes himself (and undoubtedly others) had a similar experi- 9 ence. In his address to the Roentgen Society in 1897 Silvanus Thompson sil reported that, "Crookes, after experimeting with his tube and developing some photographic plates, noted on his pictures marks that Y corresponded to his fingers; thinking the plites were defective, he returned them to the manufacturer with some strong remarks!"

~~-483 X RAY ORADICALDIAGNOSIS 4~~

~~a t I 9~~

~~18,1(' o u~~

~~1 e.~~

. Lenard and Roentgen Between 1890 and 1895 many people studied the cathoderays generated in the Crookes tube; however, most of theirexperiments were frustrating to the investigators because studies of cathoderays

(known todaias beta raysor electrons) could only be conducted . , Within the cathode-ray tube itself. Thena German scientist, Philipp Lenard, developed a tube which hada window covered by aluminum foil. He found that the cathode rays penetrated thewindow and that their effect could be observed on photographic platesseveral inches away. The rays could,also be seen faintly emerging through the win- ' dow if the room were dark and'the glass portion of thetube had a lightproof cover on it. In the 1800s many scientists were interested in studyingcathode rays. One of them was Wilhelm Conrad Roentgen, the director ofihe Physical Institute at the University of W Urzburg. On Friday, November 8, 1895, Roentgenwas beginning to conduct an experiment similar to the early ones of Goodspeedan the later ones of Lenard. As in Goodspeed's experiments, Roentgenu ed an ordinary Crookes tube connected to an induction coil. As in Lerd's research,the roomwas dark But unlike Lenard's ex entconditions, the Crookes tube had no aluminum window and was'cornpletely surroundedwith thick cardboard (Roentgen's earlier experiments with theCrookes tube had "t been done with the glass uncoveredso he could watch what was going 9n inside the tube.) It so happened.that in the darkened laboratory Roentgen's newly acquired barium platinocyanide screen(which could `detect the presence of cathode rays outside of the tube)was on a,-- LAirthelm Conrad Roentgen counter.several feet away: One of Roentgen's biographers-, Otto 11815 1,423) Glasser, has described What happened whenWilhelm Roentgen Atm-PIE atoColleQe R adilPV67 activated the lightproof tube ina pitch black room: **Suddenly, about ayard from the tube, [Roentgen)saw a weak light thai shimmered on a little bench he knew was located nearby' It was as though a ray of light or a . . faint spark from the induction roil had been reflected lay a n-qrror Not believing this .possible, he passed an other series df discharges through-the tube, andagain the same fhiorescence appeared, this time looking like 'faint green clouds Highly excited, Roentgen lit a match ancl'ie his"eat surprise discovered that the source of ttethyskentrus fight was the little bari m platinocyanide scfeen lying on,the beech Herepe ed the experiment main and again, each time movinge little screen farther away from the tube and each time getting the same result. Roentgen's' screen lit up even though thercathoderays could not escape from the tube because it had no alumirtm window. If the

184 ONLY ONE SCIENCE cathode rays could not have illuminated the screen 6 to 7 feet from the tithe*, what had caused the screen to glow? Clearly what Roentgen liqd seen was not the result of visible light or of cathode rayshe had discovered a new kind of radiation. Because at first Roentgen did not understand what the rays were, he called them "X",rays because "X" symbolizes the unknown. To record what he had seen that afternoon, Roentgen placeda

photographic plate in the path of therays and captured proof' of his 1 observation. Subsequently, he put a variety of substances between the tube and the photographic plates. He found that the,exposure of the plates changed when he used different materials and that theexposure also varied with the thicknesses of the materials. Roentgensaw that if a hand,were held before the fluoreScent screen, bones made a dark shadow nd the surrounding soft tissues produced onlyvague outlines. December 28, 1895, Roentgen submitted a manuscript d. ibing his findings to the Physical-Medical Society of Wiirzburg. Sobn Roentgen's' discovery beCame front page newsround the world. Scientists began clamoring for commercially available Crookes tubes, and when the supply ran out, investigators began to make their owri. Within'aie firstygar after Roentgen presentehis findings, more than 1,000-scientific papers anti 50 books' were published dealing with X rays. For his pioneering work-Roentgen was awarded.in 1901 the first Nobel Prize shown below.

~~ff, Stit*Mg D.~~

~~The f bel liar art antes/ 10IA ribeim C On rad RIM r1 I Q01~~

~~185 X RAYS FOR MEDICAL DIAGNOSIS~~

4 130 By early January 1896, Roentgen's pioneering paperwas in the hands of scientists throughout Europe. Theywere able to reproduce his results quickly. Late in January investigators in the United States followed suit. Then, at almost identical times, Arthur Wright of Yale University and John Trowbridge of Harvard University report&I WO- cessful experiments, which..were followedsoon thereafter by dozens of reports of others. Within a Very short time there\ vtras widespread use of X rays in medicine. They were used to locate foreign objects, toexam- ine broken bones, and to do procedures that are.of ten regardedas modern developments. In fact, at the 1896 meeting of the Association of American Physicians, reportg' were presented on theuse of X2ray plates to study the blood vessels in an amputated hand, to examinea p tientligesophagu5 following a swallow of nitrate of bismuth (a liquid, earlti k-rau ;4/0t0,4tatilt "lae //tr.bit t.jrrnrr! RootNot ich can be seen by using X rays), and to locate an aorticaurysm (a Hi_Lilt 1,!.44.5 itN th'attal dangerous abnormal ballooningof the largest artery in the y). that this I, the hatt41,4 Roolt ttt, The first

1 01 in 1897 and is shown on page 178. The subject in the picturewas . still fully clothed; and her necklace, bracelet, rings, hat pin, 'arid high-buttonshoes with nails in the heelsare clearly visible. In making this particular sicture, the subjectwas exposed to X rays for 30 minutes.

Hazards Recognized While researchers in Europe and the United Stateswere finding andi recounting multiple beneficial uses f9r Xrays, some disconcerting stories began to be circulated. Evidence was accumulating that, as the rays passed through human tissue, they could cause damage. It was ) not long before reports of serious harm from X rays began to appear. The Electrical Review reported one such case in August of 1896. A recent graduate of Columbia College, H. D. Hawks, had been working for about 4 hours a day around' an unusually powerful X-ray machine. After 4 days, he had to leave work beciuse his hands began tp swell and to appear very sunburned. Two weeks later, all of the skin came off of his hands, his fingernails stopped growing, and the hair on the sides of his head began to fall out. The hair at his temples disappeared because he routinely placed his Bead near the tube to demonstrate his jawbone to visitors who were curious about X rays. Hiseyes were very bloOdshot, and his visionwas impaired:After being treated by physicians for the severe burns and recovering somewAat, Hawks went back to work on the X-ray equipment. Following.the trial of a number of unsuccessful methods, he discoyered he could protect his hands from the X rays by covering them with tinfoil.

~~186 QNLY ONE SCIENCE There were other reports of damageus by X rays during 189,6. One equipment manufacturerG. A. Frei, prpposed that itwas not the X rays that were the cause of the problems:~~

Many physicians bring forth the argument that the, application of the X rays might prove dangerous to their patients, that here a foot had to be amputated, there someone's fingernails dropped off, another had Isore of three months' standing, etc. Such argumentscan be met with the above fact that the X-rays are pot the direct cause of the trouble, and withal/ his fact established, remedies could undoubtedly belound to reduce, if not entirely elithinate, the effect on the skin when Ire coils are to be employed Ftei claimed that it was the inductiocoil rafher than the X rays that made the problem. At the time, hev is selling static machines to pro- del duce the high iipltage electric Potential necessary to make Xrays. . were-the sa& machines usedby salespeople to demO'n-strate the mess of lightning rods. , . * . physicist at-the General Electric company;Elihu Thomsoni,,,had, ped an induction coil especially for K-rayuse thit could be used d of the static machines made by Fief. Thomsonset out to prove 'that X rays not the induction coil5 Caused the damage by deliberately malting crucial exptriments on himself wits rhe.intention of publish-,- ing the results. He askedhimself what part of his body lie could best afford to lose and decided it was the last f,bint left.little finger. He repeatedly exposed the fingertip fp-an-X.:ray tube excited bya static machine, shielding the rest of his hand with glass. Eleven days after the exposure, the skin on the back of his finger had begun feeblister and was "red, swollen, and painful to touch." A month-al-1d a halt later his finger was still sore. Thomson had proven that the burnswere xadirect , result of the rays themselves and the induction coil.was not the,cause. A year later, Thomson performed another experiment. He exposeda different finger to X rays for only a short time each day for several days' and demonstrated that the harmful effects were cumulative. By1898 Thomson's repOrtslvvinced most people- worling with Xrays to use lead shielding and to'ear protective gloves His work apparently,was not convincing enough for at least one of Thomas Edison's employees.

Thomas Edzsori Roentgen's discovery stimulated Thomas'Edis6,Q6 curiosity, and in February of 1#96 the New York World quote Edisonas saying., - . "When I have a,sufficiently powerful engine I'm sure thatt ere will be . no question of obtaining a good photograph ofta. man's ha r... I expect

~~\ . 187 X RAYS Fl MEDICAL DIAGNOSIS~~

~~A * r Ic~~

promotional photograph taken durntg Thomas Edison s attempt to take as X-ray pizotovalphsof a human bralI to have my dynamo in operation by Tuesday(February 4) at the latest." Edison was also fascinatedwith the idea of his being the first person to "photograph" a living human brain. OAFebruary 5,-he was challenged by a telegram El-0m William . Randolph Hearst, then pub- Usher of the New Yorkjournah

WILL YOU AS AN ESPECIAL FAVORTO'THE JOURNAL UNDERTAKE TO MAKE CATHODOGRAPHOF HUMAN BRAIN KINDLY TELEGRAPHANSWER AT.OUR EXPENSEr A number of repc4ters and onlookerswere Ittracted to West 'Orange, New Jersey, where Edisonhad announced he was going to ) photograph the living human brain (Figure,D).4n assistant recalled, "For three weeks, More thin tytntinewspaperreporters were sta.-. tioned A the Laboratory, the Work goingon nights, days, and Sun- daYs." The Electrical Review said, "Hehad the 'boys,' as he calls his assis- tants, working most of the time. Whert Saturdaynight came, Edison had been working steadily for'thebetter kart of 70010purs,so he- went home to rest on Sunday. Bright and earlyMonday morning he appeared at the laboratory &chipperas a lark." The Electrical I/torld rep4ted that "Edison himselfhas been having a severe attack of the Roentgen maniaK' It also told of Edison'susing the Music from a loud hand organ to keep his staff awake duringthe long liouis they were asked to work.

~~188 ONLY ONE SCIENCE A~~

~~- A Ye. A~~

~~"N.~~

(fj Jr1,1 But the scheduled photographing of the brain did not tale place scr',/t' lain,a1()/1/ C1 Ear Cv4 jer,ey on February 8. Edison explained by saying he was working on some ,,,,,,,t;vbvd special tubes for the experiment. By February 11, Edison still had not E R kt n taken a photograph of a human brain, but even that delay wa.i regarded as important by the Times, which reported that the tubes "were not ready.,He spent a busy day, however, in making experiments designed to bring out the properties of X rays." \It was not lorQbefore almost everyonin the countryas following Edison's progress, and thVe was daily reporting of the activities at t his laboratory. By February 12, reporters were getting restless. During one experiment they watched a photographic plate exposed for a full hour, and when it was developed nothing could be seen but a murki- ness and a CurveclIfirke which Edison could not explain. Edison's apol- I

~~189 X RAYS FOR MEDICAL DIAGNOSIS~~

~~19'4 I~~

ogy was quoted in the New York Times on February 13: "Aman making experiments may count himself lucky if.he has successful results inten out of every hundred experiments which he makes. At thesame time, e each negative result obtained underproper conditions closes up some avenue along which no further experiments are fteded." There were more delays and more statements. On February 23, Edison was recounting the misfortunes he had had. "My tubes have burst, I have not been able to get as high a vacuumas desired, and I have had to substitute two Leyden jars for the ondensers which I have used heretofore... I tried [a] German glass [tube.] ...andat first got good results... Afterward it failed to act altogether " The show wenton, but 1 the audience began to disappear. So far as is known, Edison never did succeed in producing that promised X-ray plate of a human brain. He did, however,manage to have work of some importance goingon in his laboratory while he was

.., out front performing for visitors. He made improvements in the Crookes tube and became known as the inventor of the fluoroscope In fact, Edison did not actually invent the fluoroscope; he did, hoVvever, make several contributions to its evolution He developeda fluoroscope which worked like an ordinary stereoscope which allowed both eyes to focus on the screen at the same time. He coated thescreen of his -fluoroscope with calcium tungstate to make thescreen much more sensitive than one coated with'barium platinocyanide: , - Edison's most important contribution to fluoroscopy, however, may have been economic. He turned the manufature of the fluoro- qcope over to a company which made it avai able sat a relatively low price;.as a result, the fluoroscope becamecon mercially available as early as March 1896. 4...-1/ Not long after Edison'sinsuccessful attempt to photograph the human brain, he became interestedin fluorescent light bulbs. He began using modified Crookes tubes as light bulbs, calling them "fluorescent lamps." According to the account of E. R. N. Grigg, theresults were disastrous: "I [raid Edison] started in to makea number of these lamps, .but I soon folind that the X ray had affected poisonouslymy assistant, ...--.. Mr. Daily, so that his hair came out and his flesh commencedto

,-. ulcerate. I then concluded it would pot do, and that it wouldnot be a very popular kind of light, so I dropped it." Clarence Daily, who was Edison's glass blower, died in 1904 before reaching his 40th birthday, It ispost likely that hewas the first fatality from X-ray "poising" in the United States.

~~190 ONLY ONE SCIENCE. s. t~~

~~s - .- F~~

2' Evolution of a New A logical consequence of the rapidly growing number of uses of Specialty X rays in medicine was the emergence of a small group of physician , who were interested in using X rays in their practice. Ata meeting of the American Roentgen Ray Society in December 1908, Percy Brown said that, 10 years before, he had heard tharany high school boy, giv a smatterittglknoidge of human anatomy, should be_able to ' X-ray pictures,' to borrow the expression used." He went on to say that in the early days all that was neededlo take X rays "was a hand to start the machine and thento stop it. Aprocess purely mechanical; no brains necessary." However, he pointed out that in 1908 physicians in isolated communities had to struggle to determine the position of fracture fragments-with a complete disregard for certain fundamental laws of optics or in blissful ignorance of the hazards of improperly used X rays, and in many cases the patient regarded the cure as worse than the disease. iknd t4en there were still tht dangers from frequently missed diagnotes. One physician attending the meeting explained his experience with X rays:

I was one of those who early tried:the roentgen raysI found, before the lapse of many years, that if I was to- do as good X-ray work as waSbeing don't by others, I ° could not practice surgery. I had no time to develop my plates except at night, when Irequired sleep. The demands of my surgical practice were too exacting to permit me to do justice to the X-ray work' I Some physicians in similar situations also chose to refer their work with X rays to other physicians who shifted the focus of their work to become "radiologists" or "roentgenologists " By 1910, 27 percent of the members of the American Roentgen Ray Society who responded to a questionnaire were involved in "Practicing roentgenology as an absolute specialty," and 90 percent of the members.responding thought that the derhand for X-ray work was growing. A very interesting finding from this same survey Was th t there was a considerable division of interests already developing a ng those who practiced radiology Forty-ninepercent of those who responded considered their practice' to be primarily diagnostic, and 19 percent categorized their use of X rays as only for therapy. Only 30 percent of the members of that Society responded that they.were using the rays for both therapy and.diagnosis. (Two perCent engaged in nei- " ther diagnosIic nor therapeutic use of )(rays.) It is noteworthy that as recently as the 1950s and 1960s radiologists and medical educators were arguing whethei or not radiology

~~191 .X RAYS FOR MEDICAL DIAGNOSIS ,~~

19S should be divided intVwo independent specialtiesofteconcerned with radiation used to diagnose medical problemsand the other specialty dealing with the use of radiationas a form of treatment or therapy.

~~DIAGNOSTIC RADIOLOGY~~

With X rays it was not difficult to localizecertain foreign substances in the Aman body, such as a bullet in the handor foot. X-ray studies of bone fractures also werecommon and very useful. But there 1- were radiologists who wanted to extend their investigations into other areas of the body, such as the brain, the gastrointestinal tract, and the cardiovascular system.

~~This early X-ray photograph shows the distribution of buck shot in a human hand fromArne, Wan PIO (Me of Radiology~~

192 ONLY ONE SCIENCE The Gastrointestinal Physicians who treated stomach and iptestinVisorders at first Tract found'no reasonto be excited about X rays. Th stomach and the small intestine produced almost no detectable X-ray hadows. The research-' el.'s were aware that they would have to devise echniques and identify signs or shado-ws which had diagnostic significance. Many radiologists were using the method which they later called "retrospectroscopy.". This-method required that the radiologists be present when the patient was qaminecat autopsy or had surgery, to learn the diagnosis, and 4), that they re- examine the Patient's X-ray photographs to look flor previously unsuspected diagnostic signs which could also be found in -\; other patients with the same condition. ° The key to X rays' usefulness for the gastroenterologist was the' discovery that if a substance opaque to the X rays cok,ild be swarlbed or introdi:iced into the bowel by enema, the intestinal tract could be "visualized." In faciNs early as 1896 W alter Cannon expen1mented with a goose and "made for it a boxso arranged that the long neck reached up thnibugh the cove. A high cardboard collar was then :attached to the-top of the box in such a way that it could be closed in front and surrounding the goose's neck. Thus the goose, with the appearance of using the most stylish neckwear, presentecko the fluorescent screen a very satisfactory extent of esophagus." At a professional meeting hat same year, the phenomenon of swallowing was demonstrated informally by having the gocise swallowcapsules con- taing bismuth subnitrate. This probably was the first public use of \\alter B 'Cannon .X rays toslembnstrate the movements of the digestiv' e tract. (1871.104i) It had been kpown for some time that extreme emotions could affect the digestive, process When Cannon was still a freshman at Ha'rvacci Medical School, he was using X rays to observe the repetitive Contractions of the stomach of a cat to which he had fed a ryieal of bread mixed with bismuth subnitrate. While he was watching, the cat "suddenly changed from her peaceful sleepiness, began to breathe quickly,and struggled to get loose. As soon astthe changelook place, the movements in the stomach entirely disappeared." He continued his obse'rvations.ond calmed the cat. As the cat relaxed and began to'purr, the rhythmic movements in the stomach began once again. Cannon also found that if he made the cat uncomfortable by holding her mouth closed and coverjag her nostrils to keep her from breathing, the movement of the stomach would stop When the cat was allowed to breathe normally, the kstricttivities reappeared These studies done long ago are still of value to modern radiologists because these studies help 4n

~~11 X RAYS FQR MEDICAL DIAGNOSIS~~

0 distinguishing disorders of the stomach whichare related to anxiety from those which are organic. As early as.1897 Charles Lester LeonardreportedThat he lladdiag- nosed a case of drooping of thestomach by introducingan emulsion of bismuth into a patient's stomach. said that thestomach had drooped so low that part orirwas visiblethrough the pelvic bones. At a medical meeting that same year, he explaintd that"by filling the hollow organs with opaque liquids, [such]as emulsions of bismuth in thestomach..their exact area can be readily clgter:mined." In the years that followed,many other physicians used bismuth as a contrast medium to study the gastrointestinaltract. About that time the.bismuth meals" began to be replaced barium sulfate ones. European radiologists had introduced the use ofbarium for their studies without any harmful effects, exceptwhen they used soluble salts of barium rather than b4rium sulfate. Bariumwas much cheaper than bismuth. Furthermore, when World Wartbroke out; American radiologists could no longer obtain bismuth fromEurope, and barium gecarne used almost universally.Barium sulfate is still one of thecon- trast mediain general use by radiblogiks. 7 During the meeting of the American RoentgenaaySociety in 1908, papers were read .on "X -ray;tidence in GastriC Ulcer," "The Roentgen Rays as an Aid in the gnosis of Carcinoma of the Stomach," and "The Roentgenographic Studyof Motion in the Nils-, eera " X rays were growing,in importanceto the diagnosis of disorders of the digestive system..

One remarkable radiologist specializingin the study of the gastrointestinal -tract was Louis Gregory Cole. He watknown as a maverick, but he also had a reputation of being extremelmeticu- lous about every detail when he was mikin0C2ay photographs. The radiographs he produced were known fortheir excgptional quality. He was described by some people as a "crankon the subject of immobil- ity" while X-ray pictureswere being taken. Cole not only insisted that patients be kept perfectly still (whichno doubt contributed greatly to the clarity of the films) but he, like otherradiologists of the'era, had a LewisC Cole ,collection of different X-ray tubes\l. He knewthe characteristics and. (1874-1°541 suitability of each one for different procedures.Henri Hulk wrote

about how a radiologist regarded histubes: . 1,1 te r1.1 11 4,17r 11.,f 1.4, .1, R He rests; pets, punishes, and smashes tubesas the fetishist his idols He learns tooknalw them allby name, their temper and their capabilitiTs- He hasa number of themthe more the better He has trained tubes,trick

~~194 ONLY ONE SCIENCE~~

19j tubes, high-spirited, high-bred tubes, as well as gentle,' steady tubes... A flashy tube is worse thin an hysteri- 1` cal woman, it is incurably useless 4 The rrecdinitiontby Cole1 and others of the characteristics ofthe individual tubes ave an impetus to the development of improvements in

X-ray ub . In 19 4 while still in his thirtiesColetlooked for sign's of stomach cancers or ulcers on the X-ray films of 566 pati4nts and made a number of correct positive diagnoses. Thirty-three of the patients on whom Cole made a negative diagnosis "presented sufficiently severe symp- toms to justify surgi,Q1exploration", 23 different surgeons ()mated on 'them. Remarkably, in all cases, sArge ricrevealed that Cole's negative diagnoses had been correct. Cole hail a so proven how invahtablFa diagnostic tool X rays could be to those who treated stomach and intestinal diseases. The revolutionary effect of X rays on diagnosis of diseases and disorders of the gastrointestinal system has continued from these early discoveries to the present. Improvements in equipmentand technique4 have made it possible for physicians to diagnose and locgte ulcers, cancers, anatomical abnormalities; trauma, circulatory problems, and a hosy,of other conditions related to the digestive system. ,, \

Another part of the body that held great interest for early (roentgenologistswas the brain. Edison was not the onlyoneinterested in photographing the living human brain. In the early 1900s, George Edward Pfahler became curious as to whether or not he could take an X-ray picture of a tumor in the brain The problem he faced was that brain tumors are composed of soft tissues that differ only slightly in X-ray opacity, if at all, from normal brain tissues or from the surrounding fluids of the brain and spinal cord. Thus, most of the X-ray photographs of thi brain examined by Pfahler did not allow the location of4 tumor or other disease in the brain Pfahler discussed-the...../.. matter with other physicians and scientists, including Arthur Willis Goodspeed, and found that there had been one distinguishable tumor demonstrated on an X-ray picture in 1899. Ifs presence had been confirmed orf- autopsy. . --% Pfahler continued thinking about using,X rays to locate tumors in the brain In 1901 he found a 32-year-old laundress who was willing to participate in his studies. Her right arm was paralyzedkand she had ter- rible,headqches and other ymptoms which suggested a large brain

~~195 X RAYS FOR MEDICAL DIAGNOSIS~~

20 L tumor. The laundress was placed in front ofan X-ray tube, and the 'filmwas exposed for 4 minutes%The plate which hadbeen placed 18 inches from the tube on the other side of the patient's skull "showed,. Iood detail of all thestructures that might be expected ina normal- ' brain; however, therewas a large shadow lying between the coronal suture and the posterior meningeal artery." The shadowwas shaped like a dumbbell, but was this thetumoror just an artifact (an artificial` characteristic introduced by the technology)?For comparison, Pfahler took an X-ray picture of anotherperson of similar size who did not have a brain tumor. Thatifilm didnot show a shadow comparable to the one seen in the launaress's x-ray Plate. The surgeons decided to operateon the patient; when they ctid, the tumor was located in theexact spot where Pfahler had said it would be, but he was not pleased.Thetumor the surgeons removed was much smaller than the one showrion his X-ray film, and it was not shaped like a dumbbell. "If I have demonstrated the tumor, then only, half of the tumor has been removed,"he said as the surgeonswere completing their operation. Unfortunately, the seriouslylil patientsur- 4- vive4 the operation for onlya few hours. Pfahler attended her autopsy which revealed "a large remainingsubcortical portion of the tumor. corresponding to the remainder andless definite part of the shadosk, " Because of this initial acciirate diagnosis,Pfahler was encouraged to perform many more eriments, primarily using the heads of cadavers. He improve theality of his radiographs and gained confidence in his diagndses forcerin types of tumors. His later studies, however, showed that the great majority oftumors did not cast a shadow that would be visibleon an X-ray plate. Therefore, other means were needed to identify these kinds oftumors. One wAy in which the diagnosis ofcertain kinds of brAin tumors was made pos'sible was by deliberatelx introducing' airinto the ventricles (the fluid-filled cavities of the /6'1 brain). Air provides a contrast to the soft tissues of the brain when viewedon radiographs, but it took Geor.te (1674 i05-) several years before radiologists beganusing this technique. I 1. .11 i1ndre A machinist living in New York Citywas the first patient in A whom it was reported that parts of the of, it, "via, brain had -been outlined because I FL411 811,11,1,1 the ventricles became filled with air. In 1912he was hit by a trolleycar H t and received an obvious head injury. liV ,;)..1.1 St1 ottr Thevachinist was taken to C Harlem Hospital where theradiologist, WilliamH Stewart, used X- ray films to diagnose a fracture of his skUll. Duringthe next 2 or 3 weeks, the man's conditiongrew, worse, and a second set of X-ray pho-

~~196 'ONLY ONE SCIENCE 4 ti~~

tograph§ were taken. "To my surprise," Dr. Stewart reported, "I found we were dealing with a condition different thari on the former exami- nation. From the shape, location, aridcfurs of those Shadows,144eir varying' and character simulating gas in thein'testines, I was led to conclude that we were dealing with a case ofIracture of ths.skull complicated by 'distended ceresbral.ventricles (inflated) with air or ges." One phYsicianEugene W. Caldwell, to Whom Stewart showed the' X-ray pictures concurred that they were seeing pictures of airin the e. ventricles, but others doubted this diagnosis: After an operation on the machinist, the surgeon reported that "two or three quick 'spits of air and fluid" were fou'ild when he entered the ca1vity of the brain. The 'surgeon also found bubbles of air in the spinal fluid.he surgery, however, had oome too late. During the autopsy of the patieAa few .days late when the brain was submerged in a tub of water aM.Ihe ventriclesere opened, air bubbled out. Apparently, in the time between wh n the first andthe second group of X-ray photographs"' were taken, thepatient hadtad a sneezing attack. The sneezing forced air through the fractUre in the frontal sinus into the vergricles. This remarkable cagtwas widely reported and discussed.\ It might be expected that news of the case of themachinist would' have prompted radiologists to begin Introducing air intothe ventricles dvliberately for diagnostic purposes, but that did not happen for some time. Duringlextseveral years, many physicians undoubtedly saw sharp outlines of the brain because of air accidentallyintroduced into the ventricles, but no-one took the creative step from accidental to de- liberate introduction of air into the brain. This delay is even more surprising when one considers that the radiologists an neurosurgeons who must hai/e viewed these plates were the same people who were Most in need of a procedure to enhance visualization of the brain The person who finallraid inject airtd visualize the brainwas a surgical resident atthe Johns Hopkins Hospitat For some time Walter E. Dandy had been izterested in methods of diagnosing brain tumors: At'the time only tumors whkh contained sialcium could be seen on an ordinary X-ray photograph Calcium is opaque to X rays, but the pro- .- cess of tumor calcification rarely occurs to the extent that it would have been visible to the X-ray pictures of that era 4 Dandy and George Heuerhisas4oCiate at Harvard, reviewed \ '.-X-ray photographs of 100 patients n whom the existence of brain tumors had been demonstrated by surgery. Only sixbf these had a shadow that might have indicated the tumor before the opeiation. In another nine cases, the tumors were obvious because`they distorted the

~~197 X RAYS FOR MEDICAL DIAGNOSIS~~

* ,cavities.of the brain The remaining85 tumors could not beseen on X-ray photographs. To obtain betterdiagnosti6eccuracy in locating brain tumors, Dan and Heuer decided that, if they'couldfill the yen,- tricles with some sstance'which 'Would producea shadow, they could find a numb r of the othertumors. Many tumors change the configuration of the walls of the ventricles.If a change in their shape could be detected, it would b%a clue that a tumor was present. Dandy and Heuer decided that whatever-theyinjected had to be nontoxic and readily absorbed and excreted bythe body. They tried many substances (including salts of bismuth, potassium,and thorium) on laboratory animals, but always thesubstance caused fatal injury to the brain Eventually; they decided itwas unlikely that they could find a substance safe enough to inject intothe brain. They were bOth unaware of any of the several published reports aboutair accidentally introdupd into the ventricles

Dandy's insight came ina very unusual way through his surgical chief, William S. Halsted.It seems that when Halsted andhisstaff were reviewing X-.ray photographs that showedbubbles of gas in the intestines, the bubbleswere recognizable even if a,layei of bonewere between them and the X-ray plate. Thiscircumstance sometimes prompted humorous remarks fromHalsted. In his jokes, he often- referred to the way in which intestinalgas had the ability "to perforate a bone " Later, when Dandy wroteabout his moment of insight, he said that his "attentionwas drawn o [the) practical possibilities [ofintro: ducing gas) into the brain ad occurred to Dandy that, if the stomach or intestines coulct beseen more clearly when they had aifin them than after they had been filledby a bismuth meal, the radiographic properties of air might bejust what he needed ro solve his ( problem. , One might think thata procedure so unusual-would be tried out first on laboratory animals; however,no animal experiments were'- reported By hilY of 1918, Dandy hadpublished an account of the first aniuman cases in which air hack been introducedinto the ventricles for diagnosticpurposes. He found that "air and water ina ventricle behaved exactly as they would ina doped flask. Following.any change in position, fluid gravitates,to themost dependent paft And the air rises to the top." Dandy was thus,ablb to Positionan air bubble anywhere in the ventricular system he wanted it by positioningthe patient's head in different Ways. His experiments led theway to many improvements in studying the brain and spinal cord,and Dandy did pioneering work using ail- to visualize other parts Of the cerebrospina,liystem.,

~~198 ONLY ONE SCIENCE~~

fr 1 The Cardiovabodar Physicians also were fascinatwith the idea of studying the cir- System c.ulatory system of the living human ody, but it was not safe to use air, bismuth, or barium for this purpose. A nontoxic substance that was opaque to X rvs, soluble in the bloodstream, and.easily absorbed by the body was neded. The first compound which satisfied these criteria was Potassitrin iodide. In 1919 its use was first repbrted by Carlos Heuser, an Argentinian radiologist. Unfortunately, his finding received little recognition, probably beCause his account was written only in a Spanish-language journal. - This use of potassium iodide and other compounds (such as sodium iodide) in the same chemical family, had to be found all over again. At the Mayo Clinic, Earl Osborne was studyin,g the effects on the body of sodium iodide injections whichlwere used to treat patients suffering from syphilis He demonstrated that when sodium iodide wallIktrOduced into the blOodstream, it was concentrated by the kidneys and then excreted. Leonard Rowntree, who was also working at the Mayo Clinic and who was interested in diagnosing kidney diseases, knew that soditim iodide was opaque faX rays He concluded >that X-ray phOtographs might be taken of the kidneys without having the-difficulties associated with injecting substances such as barium

/i to the ureters (the tubes which connect the kidneys to the bladdv), y way of a cathetgr through the bladder. His idea was an easy °rift° because patten ready receiving sodium iodide injections' to treat syphilis, andany of those same patients volunteered to have X-ray phdtographs made. T e doctors found that by using this method they could now visualize kidneys as well as the rest of the uri- nary tract by using X rays. . . Os orne's and Rowritree's work with sodium iodide was the basis Tor new work to visualize blood vesselk3hey had noted that, when the'fluoroscope was used to watch the injection of sodiuin iodide into a vein, that vein had "the appearance of a steel wire" because of the 14 iodide. In this, instance, several of those who wePe aware of this finding immediately saw its potential value in visualizing blood vessels in the human body. A rapid series of experiments proved. them'to be correct. These are just a few Of the nuniefous examples of discoveries in the early days of diagnostic radiology which were the foundation of the very sophisticated`techniques for injecting radiopaque substances into,various parts of the cardiovascular system. Because of these key discpverigs and many subgequent ones, doctors now are able to perform diagnostic procedures called angiography. These studies involve

~~409~~

~~199 X RAYS FOR MEDICAL DIAGNOSIS~~

20-4 `71*

~~f "i~~

, .1. passing a sm4.11,plastitube (a catheter)through veins or arteries. The. catheter is guided to the desiredlocation in the body while the physi- clan watches ontra fluoroscopeto determine the exact location of the catheter's tip. When the catheteris properly porSitiAed, liquid which is opaque,to X rays is injectedinto the-catheter and thus into theblood stream. This material temporarily replacesthe blood flowing through the area and outlines its arteries,capillaries, and veins. The angiograph (an X-ray photograph of the blood vessels containing theopaque sub-, stance) gives a physician importantinformation about whetheror not there is abnormal anatomyor disease in the region under study. Ithproved fluol'oscopic technologynow allows radiologists to -;- inject a substance.visible t. on X-ray photographs into a superficial vein: In the arm. They can then observe thissubstance as it passes through' the body withthe normal blood flow..Thils,problems in certain large arteries (such as those in the neck)can be identified at relatively little risk or discomfort to thepatient.

~~An inflated bill/von-tip catheter such as the one abovemay be used to open narrowedor blocked blood vessels The lower catheter is not inflated David0 DainN~~

Using frequently can inject substances which stoivessels from abnormalbleeding. This technique is not only beneficial because major,sUrgery can sometimes beavoided but is also effective insame cases m Mich sur ery is impossible.f,

~~263 ONLY ONE SCIENCE~~

V M -r . When important arteries in the body become narrow'ed, a catheter can be inserted to the site of the prpblent-Ache blood vessel is then- dilated by blowing up a tiny balloonoat the tip of the catheter..7 When the balloon is deflated and the catheter is femoved, the bldod

. c4n once again flow through the vessel, but this time withoutpbstruc- tion. This method is'now also being used in selected cases where arteries of theeart have become narrowed. The long-term effects of this procechire are not known; but there is,evidence that the balloon- cWated vessels remain open. . -fThese new methods of blocking bleeding vessels and opening obstructed ones are the/oasis of a n.ew subspecialty in the field of radiology called interventional radiology The goal of interventional I radiology is to eliminate the need for major surgery in some patients who pieltIously had nd alternative. ti a DIAGNOSTIC X-RAY EQUIPMENT Many of the advances in diagnostic radiology would not have been possiblei4ithout improvements in X-ray equipment.

Coolidge Tube Conventional Crookes tubes had many limitations. They would no! work unles's a residue of gas were left inside them. Problems occurred because the properties of the gas in the tubechlanged when the gas was heated, and these chaiacteristics varied from day to day and from minute to minute. While studying At Massachuseits Institute of Technology in 1896, William David Coolidge did not have his fascination with X rays diminished when he had to be treated for X-ray burns, however, he did gain a great respect for the rays. He was also interested in making better electric light bulbs by developing filaments which were less brittle. He developed a process to improve the tungsten that was then being used to make the filaments; he called his new product ductile/ tungsten.

4 For months,Coolidge made efforts to improve Crookes tubes by using ductile tungSten. When an ordinary tube was activated there was an electrical potential between the cathode and the anode. This caused positive ions of gas in thectube to strike the cathode. This bombardment brought about emission of electrons whiteheft struck the

~~201 X RAYS FOR MEDICAL DIAGNOSIS~~

~~I 203, O~~

The Coolidge tube was 'Faro- duced,in 1913 It provided for the first time, the ability to determine and control radiation precisely: fromThe Trail of Gmisible Light by E.R Grige,M0 4 anode and thus generated Xrays. The problem was thRt if there were not enough gas in the tube to make ion bombardment ofthe cathode possible, the,tube did not work. Thus thep'roblem of producing a S stable tube by decreasing the amount ofgas in it did not seem to have a solution unless the electrons to hit thetarget could be supplied in . o another way. An Austrian researched, Julius Edgar Lilienfeld,perceived the sam%problem that Coolidge did. Lilienfeld solved thedilemma by buildiong a tube containing a filament which,when heated, became incandescent and produced electrons. Theseelectrons made it possible to produce X rays, even though almost all of thegas had been evacuated from the tube. Lilienfeld applied fora patent that specified that electrons from the hot filamentwere used to lower the resistance of the tubc$!, A number of other peoplewere working on making a better X-iray tube, but the tube designed by Coolidgein 1913 having a,pressure "as low as it has been possible to make" wAs thebreakthrough. In his first experiments with his new tube, he recalled, "I tjrnporarilyand unin- tentionally sacrificed my own back hair. So I didn'tlike to practice on other living subjects.For further experiments Iwas, throUgh the kind- ness of a medical friend, provided with a human leg which had outlived its usefulness." Coolidge made some initial testson the cadaver's leg with an innovative tube. His intrentiOn h4 ea highvacuum, a hot cathode, and a tungsten target. He then gave his-tube to Louis Gregory Cole, the same New York radiologist whose contributions to gastrointestinal radiology already haye been described. Cole describedCoolidge's tube

202 ONLY ONE SCIENCE as -undoubtedly.lhe most important contribution-to Roentgenology since the birth of that scienEeCole published rrtaterial which pro- claimed the adv_antages of the new tube. With the Coolidge tube radiologists could expect'an exact duplication of their previous results,, and furthermore, the tubes could be adjusted very alurately Because he was-Vnown to have such high standards for his An work, praise by Cole of such an jriv:Vion promoted the tube almost overnight. The General Electric Compabwas flooded with orders for the tube from almost all of the radiologists in the United States, end the tube became standard equipment for radiologists everywhere. In principle, the Coolidge tube was also the prototype of the electronAubes usecrin radios and.computersand for many other pur o ,until their gradual 1, replacement by transistors and solid-s a e devices There were numerous other advances in the equipment that was used by radiologists Power supplies wereAmproved, and, with4the coming of World War.1 and the subsequent'unavailability of Europea glass, the glass photographic plates which were used with the overwhelming majority of the X-ray equipment were quickly replaced by film The early film, made of cellulose nitrate, was very flammable and gave oi.,kt large quantities of poisonous gas when it burned. In 1929 more than 100 people died in a firein Cleveland caused by flammable film, an incident which helped make peopld aware of the hazard. Improved cellulbse acetate-film, which was available by then, was quickly adapted aid is still in use

Image Inteniifier /Technologiessuch as the Coolidge tube and improved X-ray films werecritical1d the pcogress made iii diagnostic radiology.for about 50 years after the discovery orX rays, radiologists had seen the inside of patients by directly examining a dimly lit fluorescent screen which had been excited by X rays after passing through the patient, The image was so dithat radiologists routinely.had to spend about 20 minutes adapting thr eyes to the dark so they could see the image on the screen. In 1928 J. deBoer, G Host;and M. C. Teves of Philips Corpora- on in the Netherlands filed for a patent on the first device that formed righter image by using the electrons from the fluoiescent screen owever; they encountered what were at the time insurmountable t chnical,difficu g-and were able to achieve only a modest laboratory model, Zworykin and G. A Morton of the Radio Corporation of

~~203 X 13.AYS-FOR MEWCAL DIAGNOSIS~~

2u. America Research Laboratories published their workin1934 on a similar tithe yvhichlherhad developed toconvert invisible infrared or ultrhviblet images to visible light. During World War IIan intense effort was mounted in-Germamt and in the UnitedStates to apply these techniques towards the development ofa tube, illuminated with infrared light, whichtould be used in sniperscopes andsnooperscopes for observing the enemy at night.

It was not until1948, however, that John W ..,Coltman,a physicist at Westinghouse Research Laboratories, Dublished thereport of his. successful development of the first image intensifier Thisdevice succeeded in converting an X-ray image intoa visible image as much as 500 times as bright as those previously obtained Modern X-ray tubes employ the basic principles of Coltman's tube, but theyprovide more than 5,000 tirtes the brightness of the early standard fluorescent screertg. The invention of the X-ray image intensifier revolutionizedthe practice of diagnostic radiology Almost immediately theintensifier wak used to observe the heart in motion. At last therewas erlough light to expose movie film, and the indispensable technique of cinentioro- graphy was born. The developinents in television technologyin the 1950s made it possible to introduceclosed circuit televisionmonitoring of fluoroscopy proceduies. Improvement in, andsimplification of, television equipment in the 1960s brought about the widespreaduse of television monitors. Then video tape recorders madeit possible to rord complete fluoroseopicprocedures in realtime and also in play- bacand slow motion for more accor di gnosis. But even in the 1960s, the quality of images obtained with image -r intensifiers was not as good as the quality of imageson radiographic film primarily because the substanceon the intensifier screen) zinc sulfide powder, was not dense enough to givea sharp image. furthermore, when light_ the screen it scattered from one grain of powder to the nex,thereby causing "smearing" of the image. Eventu- ally, a tube was developed whi h used cesium iodide instead.This substance gave a surprisingly,gbod age. Research showed that when cesium iodide was applied to coatScreen (instead of depositing as a smooth, continuous layer as expected), it depositedas a layer of materialthat could be conceived ofas millions of closely packed needles, all parallel to evil other and standingup from the sheet of V glass Each needle actkd as an optical fiber conductor of the lightprO- duced in it by the X rays. (Optical fibersare discussed in more detail in

~~204 ONLY ONE SCIENCE~~

4 the chapter, "Synthetic Fibers.") The "needles" efficiently conducted the light to the surface of the instrument without scattering it side- ways to neighboring needles. Thus, the image was much sharper than

~~ever before. ,~~

Tomography A conventional radiograph is obtained by passing a beam of X rays through the body, and capturing the images projected by the X rays on two-dimensional film. But when X rays pass through a patient, dense tissue such as bone obscures less dense tissue which may lie in front of or behind it. Thus, only structures not in line with such dense ti su es Can be visualized easily. A device was deeded which would make t possible to "photbgraph" parts of the body withOot those parts b ing obscured by the shadows of intervening tissues..,_ Today such machines are calledtomographs,and the pictures they produce are tompgrams. esPerhaps Carol Mayer, a Polish radiologist, took the first step towards developing the first tomograph. Apparently in trying to get better"X-ray pictures of a patient's heart, he moved his X-ray tube. The , movementofthe tube blurred the shadows produced.by the patient's ribs, which were the part of the body closest to the tube; but the heart, which was further from the tube, had a better defined image than the

,ribs did. He published his findings but did not apply for a paten( She work of Mayer probably waenot known to Jean Kieffer, a young man afflicted withtuberculosis.Kigfer was an X-ray techni- cian who worked in a tuberculosis sanitanum in Connecticut following his own trials with tuberculosis. While confined to bed for treatment, Jean Kieffej used the time to think of,a way to visualize-the region ofs the body, the chest, where his Own disease was concentrated. The criti- cairpart of his invention was an X-ray source placed a fixed distance filom an X-ray film. The source and thefilm could be manipulated around a p. al point or fulcrumsomeWhat like two ends of a beam on a balanceale. The device could be moved in broad arcs, shallow aus, or in complex circular, curving, or"spiral patterns. During the Jean Kieffer movement, both the X-ray tube and the X-ray source rerriained in a (1897- ) proportional relationship to the pivotal point, thereby making it possible to keep an object in the plane of the pivotal point in focus on the X-ray film. Most of the shadows of objects which were not in that.

plane were "erased." Kieffer also found that as the point of rotation_ was moved cloer to or farther from the film, different planes could be brought into focus. He applied for a patent on his "X-ray'focusing- . machine" in 1929.

205 - X RAYS FOR MEDICAL DIAGNOSIS 06 For the next severalyears kieffer tried to interest manufacturers in building 'his machine. Because he had not completed highschooLind his invention was supposedto beae solution to a problem the manufacturers themselves had.notbeen able to solve,the couldnot find anyone who would give financial backingfor his device. He must have ha-mixedfeelings when he read the headlinesdi an article in the New York Times of September29, 1936, which saiZl "New X-ray Device 'Dissects' OrFilmsMachine Makes Possible Photographs of Parts ofOrgansUnobscured by Tissue." J.Robert Andrews and Robert J. Stava ofthe Cleveland University Hospitals had perfected a tomographand demonstrMed it to thosewho had arrived early for the meeting of the American Roentgen RaySociety. Kieffer immediately left forCleveland to attend the opening.dayof the meeting, which was scheduled for atefollowing day Jean Kieffer still had an important part to play inthe history of tomography. ti Most of the radiologistsat the.Roentgen Ray Society meeting were not impressed by Andrews and Stava'sdiscovery. The films which were exhibited were considered very poor froma technical point of view. Less detail seen-ledto be visible than in standard X-rayphoto- graphs In addition to Kieffer, anotherman at the meeting had tremendous interest in Andrews and Stava'sexhibit and saw great potentialin the discovery He was SherwoodMoore, director of the EdwardMal- linckrodt Institute of Radiologyat Washington University Schoobof Mediine in St Louis He struckup a conversation with Andrews anal Kieffer,(wto also had come tOvisit the exhibit booth). Mooreparticu- larly was ffnpressed 'as he heardKieffer describe hisown ideas about a similar rachine.

Moore wanted Kieffer to help himsupervise the building of sucha machine in St. Louis, and Kieffertook a le &e of absence from hisjob to do so In a 37 letter written home to his wife,Kieffer gave a candid description ois feelings. I am so tickled at the.nachine andwhat it does' that I would lite to dancea jig! pr Moore and I almost did that yesterd4 morning wl-Rn Igot a film that he didn't think could be gottena picture of an abdominalneu- rysm! The impossible' And it'was also of very much importance because it absolutely provedthe diagnosis, which was questioned bysome men. Boy Oh Boy' I was workingon the machine when Dr Moore cameup with the film as itcame out of the dark room. 'Come see this'" he literally yelledItdefinitely showed things you wouldn't guess withregular X-ray film

~~206 ONLY ONE SCIENCE vvg 1..~~

~~0 '~~

When Dr. Moore saw the long motion working [you ,0 u knoW, he was somewhat skeptical], he said, "I'won't believe it until I see it again,' and he did. He was all smiles and shoot$ my hand and said, "Kieffer, we've got them all licked."'

COMPUTED TOMOtRAPHY oft The development of X-ray tomography was an important step in improving thb physician's ability to visualize parts of the body without their being obscured by overlying tissues; however, there were other limitations of conventional X-ray technology yet to be overcome. Ordinary X-ray films'an differentiate between air, bone, fat, and water densitiel/but nbetween those of certain sofjissues which N, have densities that diff-er from one another less tlfan 21/2 percent. Moreover, Many stkctures in the body tan be visualized only by 4, introducing contrast media such as air.or substances which are opaque to X rays. Variations within soft tissues such as the brain, liver, spleen, and pancreas are usually not distinguishable, and conventional X-ray photographs cannot be used torneasure accurately the differences between objects with similar densities. The development of the tech- IP nology of computed tomography (CT) hasLade it possible to reduce these I

~~Diagram of one type of CT Scanner~~

~~207 X RAYS FOR MEDICAL DIAGNOSIS~~

~~212 .. , .~~

s -a radiologist cart obtain clear representation of a desired slice of the bOdy withiiutthe blurred-outr anatomical detail from unwanted interveningstrpctures:--a characteristic-of conventional tOnograms.-Ina CT scanner, it array of X-ray

detectors is situated opposite the X-raytube on a-large metal ring. A patient lies on a table thatcan be positioned through the center of the ring The table is then placedo the particular area of the body tobe examined is between theX-ray source and the detectors. As X rays are produced, the'ringrotates. When the X-ray source has rotated in a complete circle,Ae scan of that slice is comple . the ,4, same time the ring is rotating, the detector translatesthe measure'

ments of the changes in the X-ray beamsas the pass through the 41 object into electrical impulses whichare then transmitted to the computer The'se impulses are quantified,inathematicallymanipulated, and then used to reconstructa tomographic image. cli Computed tomography is the "reconstruction'sof an object throua particular plane (a "slice") using Xrays and a coMputer, and the machinery developed to perform theseprocedures is referred tons a CT scanner. Another term frequently heardis compuk,rized axial tomography (CAT). The latter term ismore restrictive in meaning because it refers to computer reconstnAction of cross-sectionalimages that essentially are transverse to the long axis of thepatient, whereas "CT" refers to reconstruction of a slice through the body takenat any angle. An even knoregeneral terrs computerized reconstruction. This terminology t k may be applied to nonmlEdical as wellas medical uses of the technique I and to reconstruction in eithertwo or three dimensions. One key to the development of computedto ovaphy was finding a way to reconstrudtJe internalstructure of object from X-ray measurements made exteal to the object. As earlyas 1917 an Aus- trian mathematician, Joann Radon, published his investigations ---""=------whichultimatelywere applicable to the theory of reconstruction. But

PAK., Radon was not working on image reconstruction;he was'solving ' .... gravitational field equations. Perhaps because hisresearch was in a dif- i",6 ferent field is why his work --;----4- was not known to some of thel ioneers in computed tomographywho'had to develop theirown equaions. In1956 d neurologist, William H. Oldendorf,was the first to attempt a medical application of image reconstruction.He wanteci to find a way to make diagnostic X-rayimage of the brain whichwere oftra -""t not complicated by obstruction from the oVerlyingbones of the skull. Dtagrainf Oldendorf s Oldendorf made a small block of plasticand inserted iron and alumi- Experiment num nails into it. He then put the-block on a toy train "HO"gauge flat

~~208 ONLY ONE SCIENCE~~

~~4 213 - s~~

This photograph shows Olden- doni s expertmental apparatus of MI ham °Idea°, ' car which was pulled very slowly along ashoit track by a clock motor. .1. The whole assembly was mounted on a phonograph turntablewhic'll rotatec(at 16 times per minute. Gamma rays passed through the center of rotation of the turntable, and the plastic block was pulled el through this center of rotation. Using this apparatus Oldendorf was able to obtain an approximation to his model. A*-1- Another critical link in the development of computed tom riPhy was the work Allan M. Cormack published in the early 1960s. Th story of Counack's contribution to CT scanning began in 1956 when he was a lecturer in physics at the University of Cape Town in South Africa. In that year the staff of Gioote Schuu'r Hospital unexpectedly found itself in need of a qlifiecrphysicist to supervise the use of radioactive isotopes. Becaus Cormck was the only nuclear physicist (.. * .

209 X RAYS FOR MEDICAL DIAGNOSIS in Cap Town and thepresence of one was required by law, hewas asked to spend 11/2 days each week at the hospital to fulfill this.stipula-, tion. While working at the hOspital he becamefamiliar with the plan- ningof radiotherapy treatments andwas intrigued by the complica- ticins of calculatingthe appropriate dotes of Xrays for patier4. In his own words: , It ?occurred to me that in orderto impr.Ove treatment planning one had to kow the distribution of the attenuation coefficient ofssues of the 'body, and that this distribution had toe. found by measurements made external to the boy Itsoon occured to me that this information would be usefulfor diagnostic purposes and wqulU constitute a tomogramor series of tomograms, though I did not learn the wordtomogram ALUMINUM RING for-many years. REPRESENTING Cormackknew that the basic research THE SKULL on attenuation of X rays in homogenous substances hadbeen done nearly 60years earlier, but LUCITE when he searched the literaturefor reports about what happenedwhen REPRESENTING Xrays passed through substances which THE BRAIN were not homogenous, 114 found nothing. Bcause he did not learn until 14years later that Radon had solved the ALUMINUM DISCS dblem in 1917, Corm,acy,tartedhis research without REPRESENTING this basic information 0l TUMORS One of Cormack's firstexperiments was done ona syrnmetrical phantoma cylinder of aluminumsurrounded by a wooden ring, diagramed on left. (Phantomsare nonliving experimental substitutes for patients which can provide information, usually . quantitative, about howa particular machine will perform with a living subject.) Because the formwas syTmetrical, fewer measurements were needed than foran asymmetrical form. Gamma rays- were projected from a cobaltsource through the phantom, and a Geiger counterwas used to detect the radiation which passed through the, phantom. Thedata were processed by-laborious r hand calculations, and when theywere analyzed they agreed with cross ec Irma! Dca,

~~2ty ONLY ONE SCIENCE 4~~

Cormack's work was pujikshed in 1963 and 1964 and went almost unnoticed except by a few scffihtists. About 10 years later, and totally independently, the first clinically \useful compdted tomography equipment was developetd. It was primarily the result tf the work of Godfrey N. Hounsfield who was working at the Central Research Laboratories of Electrical Musical Instruments, Ltd: (EMI) in England. (EMI is a V major force in the world's music bus sines and the producer of the Beatles' records, among others.) Hounsfield's research hactbeen in computer techniques for pattern recognition of characters written on Business forms. With this background and his interest in information retrieval in conventional X -ray examinations, Hounsfield undertook, gamma and X-ray experiments in tomographic reconstruction. He 4 speculated about the possibility of using a computer to reconstruct a picturt-from measurements of X rays that had passed through the body at many diffirent angles. Because of the multitude of measurements that would have to be made before a picture could be recon- structed from them and the great number of simultaneous equations which would have to be solved, he knew a computer was essential. 1-rou5isfield set out to see if he could develop a method to distinguish between tissues having very similar densities. In his words, "The equipment was very much improvised." A lathe bed provided the lateral scanning movement of the gamma-ray source, and sensitive ,detectors were placed on either side of the object to be viewed, which was rotated 1° at the end of each sweep. The 28,000 measurements '' from the detector were translated into numbers and automatically recorded ontaper tape After the scan had been completed, the results were fed into the computer and processed. Hounsfield's macpine produced encouraging pictures, but it worked quite slowly. It took 9 days to obtain the data by projecting X rays through a phantom and 2 1/i hours to process the results on a computer. This method obviously involved using too much radiation and was too tedious and time-consuming to be of clinical use. The next technological advance was the installation of a more powerful X-ray source, which reduced the scanning time to 9 hours. With the help of two radiologists, James Ambrose and Louis Kreel, Hounsfield used preserved humdn brains and unpreserved animal brainy to test the equipment He and his associates were successful in reGsnstructing brain.tissuebut the crucial test was whether or not tumop could be detected. Eventually, faster and more sophisticated machines were built to scan brains of living patients, and in 1972 a woman with &suspected

211 X RAYS FOR MEDICAL DIAGNOSIS brain lesion was examined using a machine developed byHounsfield. The picture revealed a circular cyst in the woman's brain.As Houns- field put it, "From this moment on, as more patientswere being scanned, it became evidentthat the machinewas going to be sensitive enough to distinguish thedifference between normal sue.' and diseased tis- The total time elapsed from Hounsfield's idea to theproductionof a functioning clinical machinewas about 5 years. The total investment in research and development during that time was about $5million. In the less than 10 years since the invention of CT,scanning, about 100 times this amount has been invested in research anddevelopment of CT scanners by industrial firms in the United States,Europe, and Asia! 11:p

This highly advanced CT system is capable of completing an entire scan of the head or body in as little as 4 8 seconds, therebymini- main image distortion caused by patient movement General Electne Company

~~212 ONLY ONE SCIENCE imullill 5~~

This image (above) ) resulted from ; preliminary CT scan of the thorax and abdomen From views simdnr to this, physicians can decide preco'ely where they taint to make more detailed cross- sectional CT 'scans such as the one shown (below) o f the upper abdomen .01 Siemens Corporation 7.4 General Electric Company

213 X RAYS FOR MEDICAL DIAGNOSIS The first use and advantageof the CT Scannerover conventional X-ray methods was in visualizing thebrain, but its value has rapidly been proven in other regions of the body. There have been.many improvementsin CT scanning since its discovery. Th4rst EMImachine cost about $300,000 in1972; in 1980 a CT scanner with simjyrcapabilities can bepurchased forabout 120,000. However, somemore sophisticated instrumentsnow may cost up to $1,200,000, and operatingcosts fora unit have been estimated to range from $259,000to $371,000 per year. The initial estimate of th*N2orld-wide market for CTscanners was for less than two dozen ma Mines. Now thenumber of scanners inuse is well over 2,000, and new ones are beingmanufacturedeachyear. Such high costs fora single piece of.medical equipment have brought CT scanners very much into the public eye, particularlyat a' time when so much attentionis being focused on the skyrocketing costs of medical care. The rapidacceptance Of CT scanning has also promoted concern over the cost of other diagnostic technologiesand even medical technology in general., Many policy issues relatedto CT scanning have arisen: How and by whom should the needforand placement of CTscanners be determined? When isa scanner safe and efficacious? Is itscost justified? Who should pay fot the technology and its operation? Someof these issues have been resolved, andconsiderable effort is continuingat national, state, and local leIs to find answers to other relatedpolicy' questions. The field of computed tomographyis still developing. Improved resolution and sNrterscan times are being sought to examinenot only the brain but also otherparts of the -Body. Curreptly, scientists, engineers, and clinicians are trying to obtainul diagnosticpictures of the heart using CT; other problems, such.asrnotion,must still be. solved. In 1979 Allan Cormack and Go y Hounsfield wereawarded the Nobel Prize in Physiology and Medine for their contributions to the development of f2complited tomograph.Neitherman had a baCk- ground in physiology or in medicine. The invention of CT isone good ' example of the dependencelirtechnologicaldevelopmentson previous discoveries in, many, diverse scientificfields. In this case, contributions from the disciplines of 'mathematics,physics, computer science,-medicine, physiology, and engineeringwere all brought together to develop a technology which probably has had thegreatest effect on the field of radiology in the 85 years sinc ilhelni Conrad Roentgen saw..4hat mysterious glow in his darkenedlaboratory.

~~214 ONLY ONE SCIENCE~~

~~2.0 0 f~~

~~.Acknowledgements~~

In addition to those mentioned in the Preface, the following people have contributed to .the de%;elopment of the Twelfth Annual Report of the National Science Board by providing expert suggestions, expert advice, information, documentation, photographs, written material, or technical assistance. The44contributions are greatly appreciated.

Bruce R. Abell + % Albert H. Cantril Richard JFox Murray Aborn Susan Cantril D,onald Frenzen Marilyn Abrams Barbara Carroll Gordon M. Frey H. Adams, Jr. John E. Casida Shigeru Fukuda Kathy Adgate Donald Chiappenelli Denos Gans Dino Agro Elizabeth Clark James F. Gibbons Carrie C. Akin Gwen Clopion John H. Gibbons John R. Akin Curtis Coker lStephen A. Gilford 1% Margaret M Akin Cynthia Conroy Roger L Gilkeson Laura Ash Jean Converse Edward L. Ginzton James Atkins Allan Cormack Edward Glass A* ""Milo M. Backus Franklin Cosmeh Henry L Goldberg Kathy Bacon Richard Coyle Derek V. Goodwin William 0. Baker Bryce Crawford Cleve Goring Albert Barber / Barbara J. Culliton Jere E Goyan Keith C. Barrons Kent Cyrtis J Thomas Grayston Victoria Berdon John Das-vi4son Adele Greene 4 Walter E. Berdon David 0 Davis Thomas Griffin Norbert M Bikales Milton Dobrin Hans Grillmeyer Robert Biork James Donahugh LaMence D. Grouse Michele Black Laurel S. Donovan Carol Hallman Jt. Blackwell John Doppman Joyce Hamaty Etcyl H. Blair Richard Drew Milton Harris 'William A Blanpied Joan Durgin David Harrison John .A Blume James Early Wt,..Vand J. Hayes, Jr. Jaan Born Robert Eilers j. C Headley Albert Bowers Stanley F. Englebardt Hollis Hedberg 1 Mary Boyd Brian Erb Mark Hegstet 1 William Boyd Gerald Estrin Charles H. Her Robin P Brett Ronald Evens William R Hewlett ,Harvey Brooks Gai' A. Fachetti Lawrence Heyl Reynold F. Brown Patricia Failla Ira M Heyman H Joachim Burhenrre, Coralie Farlee Roger W. Heyns- Arthur W Burks - W Geary Flamm Richard N. Hill, Glenna Burns Catherine Flynn Harman Hoffman , Ann L. Burroughs William Foege Norman R. S Hollies David N Byrne Virgil Freed Geld Holton

~~215 ACKNOWLEDGEMENTS~~

4 Barbara Hopkins Harry Mayne Shirley Rosenberg George L. Hutton Robert L. Metcalf Richard Royston Herbert Hyman Michael Mikuda F. James Rutherford Edward Hymoff Thomas Milby Adaline Ryan D. M. Jacobsohn Earl R Miller James Schmiedeskamp Ursula Johnsen William Miller Aaron Seriff Hamilton Johnson 0 B. Mobley, Jr Louise W Sessions Julius Johnson Paul Morgan Nathaniel Sheppard Raymond Johnson, Russell Morgan Robert E Sheriff Paul Jolly Alexander JMorin Ray Shockley JArthur Jones Audrey Morrill William B. Shockley L. B Jones Carl Moskowitz Nancy Siegal W ter JKarplus Emil M.. Mrak ElliotSivowitch Robert Kates Robert Mulholland to Leslie Sklar James B Kendrick 'William S Murray Philip Smith William Kvr Richard P. Neil Ray F Smith* Lucinda Ki4ster Daniel Newman Sukari B Smith Charles Mite!' E John Northwood Edward A Smuckler Lisa Klein Craig Nunan Mason P Southworth Nancy Knight William H Oldendorf James Stewart Donald Krotz Geoffrey Oldham Clare Stober John Kucharski Gilbert Omenn George Streisinger Arthur Kusinski D P O'Neil Arthur M Sussman Roger 0 Lambson Jane C. Orr-- Millard R. Swingle Kenneth L. Larner Mark Ouyang Bernie Talbot Otto Larsen Ronald Overmann David Torgeron Marilyn Laurie Walter Parham Darleen,Tren.eer David Leege Mary Parramg Solomon Triebwasser Martin Lefcowitz Betty Partin Earl Ubell Donald Lerch Louise Patenge Bruce L. Umminger Franklyn Levin JK. Perrell William M. Upholt Otha W. Linton Nicholas L Petrakis, Marcia Van Etten John G. Linvill Paul D. Phillips latilVingoe Mary Ann Lloyd Theodore L. Phillips Robert Vogel John A. Lofgren Frances A. Pit lick, Rose Volkert John L. Lundberg Willa Polk Sheldon L Wagner. Richard W. Lyman Phillip W. Porter Wallace K. Waterfall" Barbara S Lynch Kenneth Prewitt William Weiss Cynthia A. Lynch Judy Price Bonnie Weissman Stuart Lynn David Prowttt Lynn Welch Paul Lyons Mary Ann Rabin Storm Whaley William D, McElroy Ludwig Rebenfeld James M. Witt Dolald H. McLaughlin Leopard A. Redecke Sylvan H Wittwer Thomas Malone Robert Wood Betty W. Maloney Robert B Rice Robert R Wright Alexander R. Margulis Phillip C. Ritterbush NancjtWrubel Herman Mark John P. Robinson Fred Weingarten Kent MarqUis Sadie R. Robinson Karl Yordy Archer Martin Joe J. Robnette, Jr. James Zwolenik L. John Martin Joseph V: Rodricks George J. Maslach Joanne R Rom

~~216 ONLY ONESCIENCE U S GOVERNMENT PRINTING OFFICE 1982 -555 -1.19~~